Skip to main content

Full text of "The Journal of Cancer Research"

See other formats


Google 



This is a digital copy of a book that was preserved for generations on Hbrary shelves before it was carefully scanned by Google as part of a project 

to make the world's books discoverable online. 

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject 

to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books 

are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover. 

Marks, notations and other maiginalia present in the original volume will appear in this file - a reminder of this book's long journey from the 

publisher to a library and finally to you. 

Usage guidelines 

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the 
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing this resource, we liave taken steps to 
prevent abuse by commercial parties, including placing technical restrictions on automated querying. 
We also ask that you: 

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for 
personal, non-commercial purposes. 

+ Refrain fivm automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine 
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the 
use of public domain materials for these purposes and may be able to help. 

+ Maintain attributionTht GoogXt "watermark" you see on each file is essential for informing people about this project and helping them find 
additional materials through Google Book Search. Please do not remove it. 

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just 
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other 
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of 
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner 
anywhere in the world. Copyright infringement liabili^ can be quite severe. 

About Google Book Search 

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers 
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web 

at |http : //books . google . com/| 



I 



B OS TON 

Medical Library 



8 THE FENWAY 



r 



■ 

\ 



.4 



THE JOURNAL 



OF 



Cancer Research 



BDITBD BT 

William H. Woglom 

AfiSOCIATE EDITOR 

Fbedebick Prims 
editorial committeb 

OF 

The American Association for Cancer Research 
Joseph C. Bloodgood Lko Loeb 

Johni Hopkins Univ«nity Waihincton UniTsnity 

H. GiDBON Wblls James Ewing 

Univanity of Chieaco Cornell Univanity 

Ebnbst E. Ttizbb 

Hmnrftrd UniTanity 



VOLUME VI 



BALTIMORE, MD. 
1021 



OCT 23 19P2 
tlBRAP^ 




CO^TTENTS 

NuMBBB 1, January, 1921 

Effect of a Reduction of Lymphocytes on the Growth Rate of Transplanted 
Spontaneous Tumors in Mice. Frederick Prime 1 

A Comparison of the Growth of Mice which Ultimately Develop Carcinoma 
with the Growth of Mice which Do Not Develop Carcinoma. T. Brails- 
ford Robertson and L. A. Ray 7 

Effect of Blood from Immune Animals upon Transplantable Tumors. Isidor 
Kross 26 

Inoculation of Sarcomatous Tumors into Negro Fowls, with Special Reference 
to the Significance of Chromatophores . Yutaka Kon and Tamotsu Fuj ii . 31 

Influence of the Lymphocyte on the Peritoneal Implantation of Sarcoma in 
Mice. Ellis Kellert 41 

Primary Spontaneous Squamous Cell Carcinomas in Mice. Studies on the 
Incidence and Inheritability of Spontaneous Tumors in Mice. Fifteenth 
Communication . Maud Slye, Harriet F . Holmes and H . Gideon Wells ... 57 

Proceedings of the American Association for Cancer Research. Thirteenth 
Annual Meeting 87 

NuMBBB 2, Apbil, 1921 

A Study of a Lipomyxosarcoma with Comments upon the Origin of the Fat 

Cell. Victor C. Jacobson 109 

Parabiosis and Tumor Growth. Isidor Kross 121 

The Protein Content of the Whole Blood and Plasma in Cancer. Ruth C. 

Theis 127 

Problems in Cancer Research. Montrose T. Burrows 131 

The Influence of Heredity in Determining Tumor Metastases. Studies in the 
Incidence and Inheritability of Spontaneous Tumors in Mice. Sixteenth 

Report. Maud Slye 139 

Proceedings of the American Association for Cancer Research. Fourteenth 
Annual Meeting 176 

Number 3, July, 1921 

Further Investigations on the Origin of Tumors in Mice. VII. Tumor Age 
and Tumor Incidence. Leo Loeb 197 

Further Investigations of Disturbances of Blood Sugar Equilibrium in Their . 
Relation to Neoplasia. G. L. Rohdenburg, O. F. Krehbiel, and A. 
Bemhard 223 

Pregnancy and Tumor Growth. Isidor Kross 245 

Is Cancer Mortality Increasing? Wendell M. Strong 261 

• • ■ 

ui 



IV CONTENTS 

Number 4, October, 1921 

Cancer and Parasite. Isidor Kross 267 

Teratomas and Their Relation to Age. H. £. Himwich 261 

An Atypical Adenoma oi the Pancreas Originating in Islet Tissue. Harry 

Goldblatt 277 

The Influence upon the Growth of Transplanted Flexner-Jobling Rat Car- 
cinoma of Hydrogen Ions and of Various Salts in Different Concentra- 
tions. Kanematsu Sugiura, Helen Miller Noyes, and K. George Falk 285 

Primary Spontaneous Tumors in the Kidney and Adrenal of Mice. Studies 
on the Incidence and Inheritability of Spontaneous Tumors in Mice. 
Seventeenth Communication. Maud Slye, Harriet F. Holmes, and H. 

Gideon Wells 305 

A Critical Investigation of the Freund-Kaminer Reaction. Louis Herly 337 

Is Cancer a Biological Phenomenon? Some Heretic Thoughts on Cancer. Mrt 

F. G. Gade 357 

The Relation of Muscular Activity to Carcinoma. A Preliminary Report. 
Ivar Sivertsen and A. W. Dahlstrom 365 






OCT 23 19?2 



EFFECT OF A REDUCTION OF LYMPHOCYTES ON 
THE GROWTH RATE OF TRANSPLANTED , 
SPONTANEOUS TUMORS IN MICE 

FREDERICK PRIME 

ffwn Columbia University, George Crocker Special Research Fund, F, C. Wood, 

Director 

Received for publication April 2, 1020 

Recently much interest has been centered upon the hypoth- 
esis that the variation in the number of the hemal lympho- 
cytes may exert an important immunal function, diminishing 
or increasing the resistance of animals to tumor inoculations. 
Experimentally an increase or decrease in these cells is most 
easily induced by x-rays, and with the improved modem tech- 
nique more accurate quantitative estimations of the rays are 
obtainable than was possible a few years ago. We owe to 
Minphy and Ellis the observation that small doses of x-ray 
repeated over consecutive days reduce the lymphocjrte coimt, 
without seriously afiEecting the health of the animal (1). The 
experiments hitherto conducted on this phase of the cancer 
problem have been largely directed to the study of the effects 
of lymphocytic alterations upon transplantable tumors, where 
growth had already been well established. Sittenfield (2), 
using the Flexner-Jobling rat carcinoma, a transplantable 
tmnor which has been observed through many generations, 
found that neither increase nor reduction of the lymphoid ele- 
ments in the blood had any influence upon either resistance or 
susceptibility to tumor growth, an observation which he con- 
firmed in a later series of experiments (3). What influence, 
if any, might be observed upon the growth of spontaneous 
tumors when transplanted into such x-rayed animals has not 
as yet been satisfactorily determined. Murphy and Morton 
(4) found that such tumors transplanted to x-rayed animals 



ram ioubhal or cancbb BxuBAaoH, tol. tx, no. 1 

« 



2 FREDERICK PRIME 

"acted much as heteroplastic tissues, retrogressing as the l3nn- 
phoid tissue regenerates," and they also assert that the resist- 
ance of animals to the implantation of their own spontaneous 
tumor is increased by stimulating the production of Isonpho- 
cytes with a;-ray (5). Chambei*s, Scott, and Russ (6) have 
also reported experiments on one primary mouse tumor only. 
They observed a lowering of an animal's resistance to implielnta- 
tion after a single large generalized dose of x-ray. Since, how- 
ever, a large proportion of spontaneous tumors either do not 
grow when transplanted into other animals or give a very small 
percentage of takes, it seemed of practical importance to test 
the other aspect of the problem and to ascertain whether resist- 
ance to implantation can be decreased by a reduction of the 
lymphocjrtes. If this were possible, it would provide an easy 
method of propagating many tumors of value in cancer research, 
which otherwise would be lost owing to our inability to trans- 
plant them successfully. 

The spontaneous tiunors used were derived chiefly from a 
special strain of "Lathrop" animals, of which this laboratory 
had a large number. They were inoculated into mice from the 
same "Lathrop" strain, as such a closely related strain offers 
a more favorable soil than stock of another breed. Forty- 
ei^t to seventy-two animals in each case were used, twenty- 
four being set aside as controls, and twenty-four to forty- 
ei^t kept for x-ray treatment. 

To prepare the animals for inoculation they were given 
seven consecutive x-ray treatments according to Murphy's 
directions for reducing the number of lymphocytes. At the 
end of this time a spontaneous tumor was removed from its 
host and a small piece, about 0.002 gram in weight, was inocu- 
lated into the right groin of both the untreated controls and the 
mice whose lymphocjrtes had been greatly reduced by x-ray. 
Before starting the x-ray treatment, as well as twenty-four hours 
after the last dose of x-ray had been given, and at weekly inter- 
vals thereafter for six weeks, the blood of the x-rayed mice as 
well as that of the controls was counted. The blood pictures 
all showed a marked fall in the lymphocyte coimt after the 



GBOWTH RATE OF TRANSPLANTED TUMORS 3 

administration of seven consecutive doses of x-ray, with a 
gradual return to around the normal point by the end of six 
weeks. The controls diiring the early part of the experiment 
all showed a marked spontaneous increase in the lymphoc3rte 
count. 

The experiments lasted ten weeks or more, no ftnimala being 
included in the last series of coimts which had not survived for 
this period, and many of the series were observed over a period 
of twenty weeks. Tlie results were quite different from those 
reported by Murphy, who used tumors which had been trans- 
planted for many generations. He records that a reduction of 
the lymphocytes in the circulating blood was attended by an 
increase in the number of takes. In our experiments there was 
found to be practically no difference at all in the inoculation 
percentage between the imtreated controls and those which 
had been given a reducing dosage of rr-ray. The lymphocyte 
coimt in many of the controls was higher at the end than at 
the beginning of the experiment, and there should, therefore, 
according to the DaFano-Minphy theory, have been a smaller 
number of takes among them; but this was not the case. 

Out of a series of 905 animals living at the end of five weeks 
which had been treated with x-ray, 72, or 7.9 per cent, showed 
tumors, and 92.1 per cent were negative; while out of 740 
controls which were living, 6 per cent showed tumors, and 94 
per cent were negative. At the end of ten weeks, out of 740 
living animals which had been treated with x-ray, 4.3 per cent 
showed tmnors, and 95.7 per cent showed no growths at all; 
and out of 615 controls, 3.7 per cent of the animals showed 
tumor growths, and 96.3 per cent were negative. A difference 
of 2 per cent in the number of takes at the end of five weeks 
between the animals whose lymphocytes had been reduced and 
the controls has no significance, and a difference of 0.6 per cent 
at the end of ten weeks shows how little evidence there is in 
favor of the theory that the lymphocyte plays any important 
part in the production of immunity in relation to transplanted 
cancer (see chart 1). If the DaFano-Murphy hypothesis that 
the lymphocyte plays an important r61e in the production of 



4 FREDERICK PRIMB 

immunity were correct, a reduction in the number of tumor 
takes should have occurred in these experiments; but, as has 
been shown, such was not the case. 

That the x-ray treatment was sufficient to reduce materially 
the number of lymphocytes in the circulating blood was clearly 
demonstrated by the blood counts; and at the end-of four weeks, 



ample time for some evidence of tumor growth to be seen, the 
blood had not yet returned to normal. Whether this condition 
applies to spontaneous tumors alone, or to transplantable 
tumors in general, will have to be demonstrated for each tumor. 
In the series reported, over forty spontaneous tumors were 
used and inoculations were made into 2100 mice. 

CONCLUSIONS 

Decreasing the circulating lymphocytes by small doses of 
x-ray does not iiender mice more susceptible to the inoculation 



GROWTH RATE OP TRANSPLANTED TUMORS 5 

of a spontaneous tumor from another mouse of the same 
strain. 

Mice with spontaneous lymphocytosis are not resistant to 
the implantation of a primary tumor from another mouse of 
the same strain. 

REFERENCES 

(1) MuRPHT, J. B., AND Elub, A. W. M. : Experiments on the r61e of lymphoid 

tissue in the resistance to experimental tuberculosis in mice. J. 
Exper. M., 20: 397, 1914. 

(2) SiTTENFiBLD, M. J. : The significance of the lymphocyte in immunity to can- 

cer. J. Cancer Res., 2: 151, 1917. 

(3) SiTTENVTELD, M. J. : Further studies on the importance of the* lymphocyte 

in cancer immunity. J. M. Res., N. S. 88: 403, 1918. 

(4) MxjBPHT, J. B., AND MoRtON, J. J. I The lymphocyte in natural and induced 

resistance to transplanted cancer. J. Exper. M., 22: 204, 1915. 

(5) MxjBPHT, J. B., AND MoBTON, J. J. I The effect of Roentgen-rays on the rate 

of growth of spontaneous tumors in mice. J. Exper. M., 22: 800, 1915. 

(6) Chambers, H., Scott, G., and Russ, S.: On the action of x-rays upon the 

transplantation of a spontaneous carcinoma of the rat. J. Path, and 
Bacteriol., 28: 384, 1919-20. 



A (COMPARISON OF THE GROWTH OF MICE WHICH 

ULTIMATELY DEVELOP CARCINOMA WITH 

THE GROWTH OF MICE WHICH DO 

NOT DEVELOP CARCINOMA 

T. BRAILSFORD ROBERTSON and L. A. RAY 

(From the Department of Physiology and Biochemiatry, University of Adelaide, 
South Atutrdlia, and the Department of Biochemistry, University of Toronioy 

Received for publication July 6, 1020 
STATEMENT OF THE PROBLEM 

Since cancer is a disease of growth, and its incidence in appre- 
ciable form is usually deferred until after the attainment of 
maturity, a very important factor in our ultimate understanding 
of the etiology of cancer must arise out of a knowledge of the 
degree to which the preceding development of the animal is 
affected by the ultimate aberration of growth. If the factors 
which determine the growth and incidence of carcinoma, for 
example, are of an essentially accidental character, such as local 
irritation plus infection, then the incidence of carcinoma in an 
individual will be decided in the main by factors not inherent 
in its habit of growth and metabolism, and the preceding 
normal development cannot be expected to foreshadow the 
impending pathological outcome. If, on the contrary, the 
incidence of carcinoma is determined by internal factors of 
growth-habit or metabolic peculiarities plus recurrent external 
factors, such as local irritation, then the underlying peculiarity 
of growth or metabolism may conceivably reveal itself in more 
or less marked departures from the average in the preceding 
normal development. 

^ The expenses of this research were in part defrayed by a grant from the 
special medical research fund of the University of Toronto. 



8 T. BRAILSFORD ROBERTSON AND L. A. RAY 

A comparison throughout the duration of life of the growth 
of animals .which ultimately develop spontaneous cancer and 
those* which do not is, therefore, fundamental to our compre- 
hension of the etiology of this disease. The fact that such a 
comparison has not hitherto been instituted is imdoubtedly 
attributable to the experimental labor which is essential to obtain 
results of any worth, to the expenditure of time which is requisite, 
and to the difficulty of an experimental technique which must 
obviously be devised to exclude all chance causes of death, 
among which must be reckoned epidemic infections. It is 
obvious that if epidemic infections were from time to time to 
claim a substantial proportion of victims among the experi- 
mental animals, it would become impossible to identify the 
individuals which, had they survived the infection, would subse- 
quently have developed carcinoma. 

BCETHODS OF RESEARCH 

For the past six years we have been engaged upon an extensive 
series of experiments of which the immediate object was the 
determination of the influence of certain glandular extracts or of 
normal food constituents in extraordinary excess upon the 
growth of otherwise normal white mice. The imusual constit- 
uent was added to an otherwise varied and abundant diet of 
barley, egg, biscuit, and lettuce. It was essential to our inquiry 
that epidemic infections should be excluded, and we were so far 
successful that mortality from these diseases, throughout the 
duration of their lives, did not amoxmt to more than 4 per cent 
of the total deaths in the males, and 1 per cent in the females. 
The animals were weighed once a week from the fourth or fifth 
week after birth imtil the thirtieth week, and thereafter once a 
fortnight. Individual records of the growth of each mouse were 
kept, and its ultimate fate recorded. All lesions manifest to the 
naked eye in postmortem examination of the organ were noted, 
and sections of formaldehyde-hardened material were prepared 
and examined microscopically. We feel confident that nearly 
all cases of carcinoma and other new growths were detected and 



GROWTH OF MICE AND CARCINOMA 9 

identified except in a very small proportion of cases in which 
postmortem examination of the tissues was rendered impossible 
by delayed observation of death and consequent decomposition. 

The details of our experimental technique and the growth 
and mortality statistics thus obtained have been fully described 
elsewhere (1). 

By segregating; in each experimental group, the records of 
those animals which ultimately developed carcinoma, we have 
been able to comjiare the growth of those animals which 
developed carcinoma with the growth of those animals which 
did not, in eleven different experimental groups, involving a 
total of 324 animals, of which 105 developed spontaneous carci- 
noma. The primary growth was usually situated in the axilla 
or the groin, and metastases occurred in the lungs. In a small 
proportion of cases in the other groups and in every case in the 
group receiving tethelin throughout the duration of life (tethelin- 
f ed males) the only new growths detectable by naked eye exami- 
nation were situated in the limg; and these, in the tethelin-fed 
group, were very small, frequently but 1 to 4 mm. in diameter. 
In the more typical cases, involving a primary growth in the 
axilla or groin, death usually occurred within one month of the 
appearance of a manifest lump. 

The diet of the "normal" animals consisted of crushed barley 
and water ad libitum, fresh leaves of lettuce twice a week, dry 
hard imsweetened biscuit ("pilot bread") once a week, and 6 
cc. of mixed white and yolk of egg to every six animals on each 
of six days of the week. The egg was employed as the vehicle 
for the administration of the various dietary additions to the 
other groups; it was generally consimied within a few minutes of 
its introduction into the cage. 

The pituitary-fed animals received one-half of an anterior lobe 
of an ox pituitary per six animals, emulsified in the daily allow- 
ance of egg. The lecithin, prepared from yolks of eggs, was 
administered in dosage correspondmg to 83 mgm. per mouse per 
day. The dosage of cholesterol was 42 mgm. per mouse per day. 
The dosage of tethelin (a lipin prepared from the anterior lobe 
of the pituitary gland) was 4 mgm. per mouse per day. In one 



10 T. BRAILSFORD ROBERTSON AND L. A, RAY 

group (tethelin-fed males) the administration of tethelin was 
continued throughout life. To one group of females ("discon- 
tinuous administration") this dose of tethelin was administered 
in three periods of one month each, namely from the end of 
the fourth week imtil the end of the eighth, from the end of the 
twenty-first until the end of the twenty-fifth, and from the end 
of the forty-second until the end of the forty-sixth week of life. 
To another group of females ("brief administration") the admin- 
istration was continued only from the end of the fourth imtil 
the end of the twelfth week, and thus ceased before or at about 
the attainment of sexual maturity. Weighings, in this latter 
group, were discontinued between the thirtieth and the seventieth 
week because, mistakenly as eyents proved, no noteworthy effects 
of the administration upon normal growth subsequent to the 
thirtieth week were anticipated. 

■ 

EXPERIMENTAL RESULTS 

The incidence of carcinoma in the various experimental classes 
and the average duration of life of the animals in each class and 
of those which developed carcinoma are enumerated in table 1.^ 
It will be observed that the percentage of incidence of carcinoma 
was substantially unaffected by the various dietaries employed, 
with the exception that the incidence was extraordinarily low 
(17 per cent) in the group of females which received the "brief 
administration" of tethelin. Traces of the same effect may be 
seen in the low incidence (25 per cent) in the group which received 
"discontinuous administration" of tethelin, the normal iQcidence 
in females beiog, for the strain employed, from 30 to 50 per cent. 
We shall see that of all the growth curves herein presented that 
for the "brief administration" tethelin group departed most 
widely from the type of growth curve generally displayed by 
animals which subsequently developed carcinoma. The high 
incidence of carcinoma iu the pituitary-fed groups may possibly 
be fortuitous, or, again, may be attributable to the pituitary 

' Owing to the cost of publication it has been necessary to omit many tables 
bearing in detail the data of the experiments. — Editor. 



GROWTH OF MICE AND CARCINOMA 



11 



tissue, or, yet again, to the fact that the diet of these animals 
differed from that of all other classes in containing a proportion 
of raw meat. 

The results of the comparison of the growth of the animals 
which ultimately developed carcinoma with the growth of the 
animals which did not are displayed graphically in Figures 1 
to 11, in which the full lines represent the growth of the animals 
which did not, and the broken lines represent the growth of the 

TABLE 1 

Incidence of carcinoma itk the various experimental classes and the duration of life 

of each 



CLA88 OP ANIMAU 



AXIMALS WHICH 
DXYBLOPBO 
CABCINOMA 



AYXBAOB DUBA- 

TION OP LIPX 
OP ALL ANIMAU 



ATBBAOB 

DUBATIOir OP 

LIPBOPTHB 

AMIMALfl WHICH 

DXTBLOPHD 

CABCnCOMA 



Males 






Normal 


per cent 

25 
29 
22 
26 
35 


dayt 

767 
792 
731 
764 
866 


data 
826 


Pituitary 


823 


Lecithin 


743 


Cholesterol 


848 


Tethelin 


958 







Females 








Normal 


1 


37 
51 
32 
50 
25 
17 


719 
704 
731 
658 
800 
695 


721 


Pituitary 


689 


Lecithin '. 


743 


• 

Cholesterol 


590 


Tethelin (discontinuous administration). 
Tethelin (brief administration) 


970 
677 







animals which did develop carcinoma. The comparison is dis- 
continued in each case at the date at which half of the carcinoma 
animals had died. 

An examination of these curves reveals a striking uniformity 
of result. In each of the eleven different experimental groups 
the carcinomatous animals were superior in weight to the non- 
carcinomatous throughout a large proportion of their lives, and 
this occurred independently of whether, at the beginning of the 
weighings at four or five weeks of age, they were superior, equal, 



12 



T. BSAILBFOBD B0BEBT80N Ain> U A. BAT 




cT 



Fig 1. CoMPABisoN of the Growth of Normal Male White Mice Which 

Ultimately Developed Carcinoma (Broken Line) with that of 

Normal Male White Mice Which Did not Develop 

Carcinoma (Full Line) 



.#•%< 



'••#: 



? 



Fio. 2. Comparison of the Growth of Normal Female White Mice Which 

Ultimatelt Developed Carcinoma (Broken Line) with that of 

Normal Female White Mice Which Did Not Develop 

Carcinoma (Full Line) 



OBOWTH OF MICE AND CABCINOMA 



13 



or inferior in weight to the non-carcinomatous animals. The 
superior growth occurred during the period of developing sexual 
maturity (adolescent growth) and was merely maintained or 
slightly diminished or even lost during later life. These facts 
are summarized in table 2. 

It will be seen that in four out of the eleven groups the animals 
which developed carcinoma were initially inferior in weight to 



TABLE 2 



Relative weighUt at different ages of animals which developed carcinoma and those 
which did not. + indicates superiority of the animals which developed cancer; 
-^ indicates their inferiority in weight, and => indicates substantial equality of 
tfie two groups 



OiAflB OF ANIMAUB 



imrxAi. 
mBxoirr (4 

TO 5 WKKK8 



10 TO 80 



80 TO 70 

WSBK8 



70 WBBXB TO 

DBATH or 

OKB BALV OF 

AjraCALB 

WHICH 

DBTKiOPXD 

CABCINOMA 



Males. 



Normal 

Pituitary. . . 

Lecithin 

Cholesterol. 
Tethelin... 



» 


+ 


- 


=» 


-h 


— 


-h 


-h 


-h 


-h 


-h 


+ 


- 


+ 


+ 






Females 



Normal 

Pituitary 

Lecithin 

Cholesterol 

Tethelin (discontinuous adminis- 
tration) 

Tethelin (brief administration) 



-h 



+ 


+ 


— 


+ 


+ 


+ 


— 


+ 


+ 


+ 


+ 


+ 


+ 


+ 


^ 


+ 


+ 


+ 



the animals which did not subsequently develop carcinoma. 
Yet even in these cases superiority of weight was ultimately 
attained, and in three out of the four groups superiority of weight 
was attained before the thirtieth week. In the remaining group 
(lecithin-fed females) the initial inferiority was very great 
(1.66 grams, or 15 per cent of the mean weight) yet this group 
attained equaUty to the non-carcinomatous animals before the 
thirtieth week, and smpassed them thereafter. 



14 



T. BRAILSFORD BOBERTSON AND L. A. SAT 




c/ 



Fig. 3. Compabison of the Growth of Pituitaby Fed Male White Mice 

Which Ultimatbly Developed Cabcinoma (Bboken Line) with 

THAT ol* Pituitaby Fed Male White Mice Which Did 

Not Develop Cabcinoma (Full Line) 



^mm 




PfTWlTMV 



? 



-r 

a 






A 



Fio. 4. Compabison of the Gbowth of Pituitaby Fed Female White Mice 

Which Ultimately Developed Cabcinoma (Bboken Line) with 

THAT OF Pituitaby Fed Female White Mice Which Did 

Not Develop Cabcinoma (Full Line) 



GROWTH OF MICE AND CARCINOMA 



15 



INTERPRETATION OF THE RESULTS 

We may infer that the animals which ultimately develop 
carcinoma foreshadow this outcome in their development during 
the earlier portions of the third or adolescent growth-cycle. 
They are distinguished from the animals which do not subse- 
quently develop carcinoma by their exceptionally energetic growth 
during the period of adolescence. Prior to this period they do 
not display any imiform characteristics of weight, and subse- 
quently to the thirtieth week the growth curve of the carci- 



.•"♦, 



'•••*•• 



•---•« 



#-•: 



d" 






T 



Fig. 5. Cobiparibon of the Gbowth of Tbthblin Fed Male White Mice 

Which Ultimatblt Developed Carcinoma (Broken Line) with 

THAT OF Tetheun Fed Malb White Micb Which Did 

Not Develop Carcinoma (Full Line) 

nomatous animals usually remains parallel to the curve for the 
non-carcinomatous animals or approaches it, the only marked 
exception to this rule being afforded by the group of cholesterol- 
fed females in which the animals which subsequently developed 
carcinoma display a late accretion of weight between the forty-^ 
fifth and the sixty-fifth week. 

The same data have afforded material for another significant 
comparison, namely, that of the growth of animals which sur- 
vived beyond the average duration of life in any given experi- 
mental group with the growth of the animals which failed to 
attain the average duration of life (2). In this comparison 



16 



T. BRAILSFORD ROBERTSON AND L. A. RAT 



it was shown that the long-lived animals are distinguished by 
relatively energetic early growth, low variability, steadiness of 
weight, and absence of late accretions of weight. The short- 
lived animals, on the contrary, display relatively deficient early 
growth, high variability, instabiUty of weight, and a marked 
tendency to acquire late accretions of weight. . 

The animals which ultimately develop carcinoma, therefore, 
display the general characteristics of the long-Uved groups of 
animals. At first sight this is not surprising, for the average 
duration of life of the animals which develop carcinoma exceeds, 



>••••, 







? 



V'AJl^ A ^ I ^ s « « 9«& A it 4 

Fio. 6. Comparison of the Gbowth of Tethelin Fed Female White Mice 

(Discontinuous Administration) Which Ultimately Developed 

Carcinoma (Broken Line) with that of Tethelin Fed 

Female White Mice (Discontinuous Administra- 

tration) Which Did Not Develop 

Carctnoma (Full Line) 

in eight of the eleven experimental groups, the average life- 
duration of the group. On closer examination, however, this 
result is not so readily intelligible as it appears. In the first 
place, the carcinomatous animals display the characteristics of 
the long-lived animals to an exaggerated degree, and, in the 
second place, only a little over two-thirds of the carcinomatous 
animals actually belonged to the long-Uved group (table 3). 
On the other hand, of course, many long-lived animals and, in 
fact, most of those which survived for the longest periods, failed 
to develop carcinoma at all. 



GROWTH OF MICE AND CARCINOMA 



17 



We cannot, therefore, conclude that these results are due simply 
to the longevity induced by the superior growth-energy of the 
carcinomatous animals passively carrying them into the age- 
zone in which carcinoma occurs. A majority of all the animals 
attain the ages at which one-third of the carcinomatous animals 
die. A smaU minority of animals exceed the age at which the 
last carcinomatous animal has died, without developing carci- 



TABLE 3 



Showing the proportion of animals which developed carcinoma which also lived for 

longer than the average duration of life 



CLAB8 OF AUnUAlA 



TOTAI<NUMBEB 
or ANIMALS 



KUMBKB or 

ANIMALS WHICH 

DBVKLOPED 

CABCINOMA 



NUMBXBOr 
ANIMAUB DB- 
VSLOPINa CAB- 
CINOMA WHICH 
ALSO WBBB 
LONG LIYXD 



Males 



Normal 

Pituitary... 

Lecithin 

Cholesterol. 
Tethelin.... 



Totals. 



32 
31 
32 
34 
23 



152 



8 
9 
7 
8 
8 



40 



6 
7 
5 
7 
7 



32 



Females 






Normal 


32 
35 
34 
32 
16 
23 


12 
18 
11 
16 
4 
4 


8 


Pituitary 


11 


Lecithin 


6 


Cholesterol 


9 


Tethelin (discontinuous administration) . 
Tethelin (brief administration) 


4 
2 






Totals 


172 


65 


40 







noma at all. Moreover, while the lowest incidence of carcinoma 
recorded in table 1 occurs in the shortest-lived group (tethelin, 
"brief administration") yet the highest incidence occurs in a 
group which lived an average of but twenty-four days longer 
(pituitary-fed females), and the next lowest incidence occurs m 
the longest-lived group of females (tethelin, "discontinuous 
administration") • 



18- 



T. BRAILSFORD ROBERTSON AND L. A. RAY 



A remarkable feature of the results is their uniformity. This 
becomes the more noteworthy when we consider the variability 
of the experimental material. No estimates of variability in 
quantitative terms have been computed, because in many cases 
the small nimiber of animals which ultimately developed carci- 
noma in any one experimental group rendered such a computation 
of very little value, Nevertheless, it may readily be perceived 
that there is no single type to which the carcinomatous animal 





? 



Fig. 7. Compabibon of ths Gbowth of Tethbun Fed Female White Mice 

(Brief Adminibtbation) Which Ultimai^lt Developed GABCiNoiiA 

(Broken Line) with that of Tetheun Fed Female White 

Mice (Bbisf Administration) Which Did Not 

Develop Cabcinoma (Full Line) 

adheres. No such simple rule can be formulated as that heavy 
animals develop carcinoma. Thus in the tethelin (''brief adminis- 
tration") group one of the animals which developed carcinoma 
weighed but 22.5 grams at seventy weeks, or .nearly 7.5 grams 
less than the average weight at the same age of the non-carci- 
nomatous animals of the same experimental group. Similar 
phenomena were observed in the other experimental groups. 
For this reason, of course, it is impossible to foretell the incidence 



OBOWTH OF MICE AND CARCINOMA 



19 



of carcinoma in any individual animal from its preceding growth. 
How then does it arise that the average weight of the animals 
which subsequently develop carcinoma is so invariably in excess 
of the average weight of non-carcinomatous animals throughout 
a large proportion of their lives? 

There would appear to be only one possible explanation, 
namely that it is superiority of growth relatively to the standard 
norm of the strain which foreshadows carcinoma, not energetic 
growth per se. The animals employed in this, as in any other 



^rm 




d" 



r 



Fig. 8. Comparison of thb Gbowth of Lecithin Fed Male White Mice 
Which Ultimately Developed Carcinoma (Broken Line) with 

THAT OF IiEblTHIN FeD MaLE WhITE MiCE WhICH DiD NoT 

Develop Carcinoma (Full Line) 

animal experiment, represented a variety of more or less widely 
differing strains, each possessed of its own standard of absolute 
and relative development. In any suflSciently large group of 
animals chosen at random from the general stock, each of the 
main strains would be represented in proportion to its frequency. 
Since the carcinomatous animals, although so variable among 
themselves, invariably displayed an average weight which for a 
large period of thfeir lives exceeded the average weight of the 
remainder, we must conclude that each of the various strains 



20 



T. BRAIL8FORD ROBERTSON AND L. A. RAY 



in the stock furnished its proportionate quota to the carcinoma- 
tous group, and in each strain the animals which ultimately 
developed carcinoma were superior in energy of growth to the 
standard norm of the strain. This fact precludes the idea that 
we may be dealing with linked size and carcinoma-inheritance, 
the carcinomatous animals representing a single strain or a small 
number of strains distinguished also by large size. The two 
factors, that of absolute size and that of tendency to develop 
carcinoma, are, on the contrary, separable. 



/•--- 



? 



T 



T 



A 



T 



Fig. 0. GoMPABiBON op the Growth of Lecithin Fed Female White Mice 

Which Ultimatelt Developed Carcinoma (Broken Line) with 

THAT OF Lecithin Fed Female White Mice Which Did Not 

Develop Carcinoma (Full Line) 

In the article dealing with the comparison of the growth of 
long-lived with that of short-lived animals, it was sought to 
interpret the results in terms of the competition between paren- 
chymatous and connective tissues during the development of 
the animals. In those animals in which the anaboUsm of paren- 
chymatous tissues is exceptionally rapid, the growth of paren- 
chymatous structures will be hastened and^ faciUtated. This 
may or may not tend to increased weight of the animal at the 
time, according to the degree of acceleration experienced and the 



GROWTH OF MICE AND CARCINOMA 21 

relative masses of parenchyma and connective tissues in the 
body of the animal. It will, however, manifestly delay, through 
previous appropriation of foodstuffs, the senescent accretion of 
connective tissues, and consequently increase the duration of 

life. 

If we apply this hypothesis to the results enumerated in this 
paper we must conclude that the animals which subsequently 
develop carcinoma are distinguished by exceptionally rapid 
anabolism and growth of parenchyma which, during adolescence, 
more than compensates for the complementary delay in con- 





(/ 



T 



FlO. 10. QOMPARISON OF THE GbOWTH OF CHOLESTEROL FeD MaLE WhITB 

Mice Which Ultimately Developed Cabcinoma (Bboken Line) 

WITH THAT OF CHOLESTEROL FeD MaLE WhITE MiCB WhICH 

Did Not Develop Carcinoma (Full Line) 

nective tissue growth. In the absence of carcinoma this would 
result in great longevity, which however, is curtailed by a patho- 
logical proliferation of parenchymatous tissue. We may suppose 
that in such animals the energy of response to some external 
stimulus, such as a local source of irritation, is so excessive that 
when the age of the animal is sufficiently advanced to have 
resulted in some weakening of the competition of other paren- 
chyma (through the parasitism of connective tissues) the effect 
is to institute a growth which appropriates nutriment to itself 
at the expense of all the adjacent tissues. 



22 



T. BRAILSFORD ROBERTSON AND L. A. RAY 



In accordance with this hypothesis we find that the incidence 
of carcinoma is lowest in that experimental group in which the 
growth curve departs most widely from the long-lived type, 
namely, in the tethelin "brief administration" group in which 
there is an extraordinary late accretion of tissue (3). The only 
other group of carcinomatous animals which reveals late accre- 
tion of weight is also the group which has the briefest duration 
of life (cholesterol-fed females, table 1). 



.— ' 



,-*— ' 



f~^ m m ' 



»,,,,,. ,#>****«*^ 



? 



II 



It 



•t 



FiQ. 11. Comparison of the Gbowth of Cholbstebol Fed Female White 
Mice Which Ultimatelt Developed Cabcinoma (Broken Line) 

WITH THAT OF CHOLESTEROL FeD FeMALE WhITB MiCE WhICH 

Did Not Develop Carctnoma (Full Line) 



The longest-lived animals in any group must, therefore, be 
those in which the speed of anabolism and energy of growth of 
the parenchyma just fall short of the tendency to respond abnor- 
mally, by the formation of neoplasms, to recurring local irritative 
stimuli. 

The comparative growth of animals which ultimately devel- 
oped neoplasms other than carcinoma will form the subject of 
subsequent communications. 



GROWTH OF MICE AND CARCINOMA 23 

SUMMARY 

1. The incidence of carcinoma in mice is foreshadowed in 
their preceding development; but, owing to the variability of 
the animals, it is not possible tp foretell the incidence of carci- 
noma in any single animal from its curve of growth. 

2. The animals which ultimately develop carcinoma are 
distinguished by relatively energetic growth during the period 
of adolescence. The lead over the other animals which is thus 
established is usually maintained throughout life, but is some- 
times lost in the later periods of life through late accretion of 
weight by the abimals which do not develop carcinoma. 

3. These results do not admit of interpretation by supposing 
that superior growth favors longevity, and thus passively carries 
the animals into the age-zone in which carcinoma occurs. A 
majority of all the animals surpass -at death the ages at which 
one-third of the carcinomatous animals die. There is no cor- 
relation between the percentage incidence of cancer in any group 
and its average duration of life; but animals displaying an aver- 
age growth curve which departs widely from the carcinomatous 
type also display a low percentage incidence of carcinoma. 

4. The results are not due to linked inheritance of size and 
tendency to develop carcinoma, since the factors of absolute 
size and development of carcinoma are separable. 

5. The results are interpreted to mean that the animals which 
ultimately develop carcinoma are those in which the anabolism 
and, therefore, the growth of parenchjrma is exceptionally rapid. 
In such animals the energy of response to local irritative stimuli 
may be so excessive as to overcome the competition of other 
tissues and initiate a new growth. 

6. The longest-lived animals in any group are, therefore, those 
in which the speed of anabolism and energy of growth of the 
parenchjona just fall short of those which lead, under the influ- 
ence of recurrent irritative stimuli, to the formation of new 
growths. 

REFERENCES 

• 

(1) RoBSBTSON, T. B., AND Rat, L. A.: Jour. Biol. Ghem., 1916, xxiv, 347, 363, 

385, 397; 1916, xxv, 635, 647; 1919, xxxvii, 377, 393, 427, 443, 455. 

(2) RoBEBTSON, T. B., AND Rat, L. A.: Jour. Biol. Chem., 1920, xlii, 71. 

(3) RoBEBTSON, T. B.. AND Rat, L. A.: Loc. cit. 



EFFECT OF BLOOD FROM IMMUNE ANIMALS UPON 

TRANSPLANTABLE TUMORS 

ISIDOR KROSS 

From Columbia Untvernty, George Crocker Special Research Fund, F, C. Wood, 

Director 

Received for publication October 22, 1920 

Even a cursory review of research in experimental cancer will 
reveal many inconsistencies, various workers having obtained 
contradictory results when attacking a problem with apparently 
similar methods. In many cases, however, patient repetition 
with careful checking up of the findings has brought to light 
factors which offer a satisfactory explanation for these puzzling 
variations. In this study an attempt was made to clear up one 
of the problems in which different workers have reached dia- 
metrically opposed conclusions. 

Jensen (1) reported the disappearance of transplanted tumors 
in mice injected with serum from rabbits treated with carcinoma. 
In a later publication (2) he stated that while he realized that 
spontaneous retrogression would explain most of his ''cures/' 
it did not account for the disappearance of several of the very 
large tumors. 

von Leyden and Blumenthal (3) reported the successful out- 
come of the treatment of a dog with a carcinoma, the diagnosis 
of the growth having been verified by microscopical examination. 
The method was as follows: Several rabbits were given a series 
of subcutaneous injections of carcinoma emulsion over a period 
of several weeks> and the serum obtained from these rabbits 
was injected subcutaneously into the dog. 

Clowes (4) reported the results of experiments with mice in 
which one half of the animals received injections of ''immune" 
serum obtained from mice whose tumors had retrogressed spon- 
taneously. The other half acted as controls and received normal 

25 



26 ISIDOB KB068 

mouse serum. Of the twenty mice in the first series, onfy one 
failed to show the beneficial effect of the immune serum, and aO 
were alire at the time of the report; while in the control series, 
five were dead and all the others had tumors exceeding in siae 
those of Ihe first series. Clowes and Baedack (5) reported the 
results of experiments in which one lot of mice was treated with 
tumor mixed with immune serum, a second with tumor wiTg«d 
with normal serum, and a third with tumor mixed with normal 
saline solution. The difference between the results obtained in 
the two latter groups was negligible. The difference between 
the results in the "immune" set and in tiie two contrcd sets, 
however, was considerable, 31.6 per cent of tumors devdqping 
in the controls and only 12.3 per cent in the immune group. On 
the basis of these findings, they justified their assumption of the 
presence in the blood of inmiune bodies antagonistic to the devd- 
opment of cazcmom&. 

Crile and Beebe (6) rqwrted a series of ten blood tnnsfusiaDs 
in dogs with an infectious Ijrmphosaicoma, resulting in cure 
in seven cases and in mariced improvement in two. In the 
remaining case, in which the treatment was a complete failure, 
it was later found that the donor was not an immune animal, 

von Dungem (7) obtained serum from rabbits in which there 
has occurred spontaneous retrogression of a sarcoma. Seven 
rabbits w&e treated by intravenous injection of this serum befcxe 
being inoculated with sarcoma. Nine other rabbits were em- 
ployed as controls. Six of the control animals developed tumors; 
in none of the tareated animals did a tumor devdop. TJhlenhuth, 
TTshHaI^ and Steffenhagen (8), however, p^ormed the same 
experim ent and found that rats treated with "inmiune" serum 
gave 94 per cent of successful takes, whereas those treated with 
normal s^rum gave 83 per cent, and those that received no treat- 
ment at all gave only 66 per cent. In another series of experi- 
ments, these authors mixed an emulsion of tumor tissue with 
immune semsm. The injection of this mixture gave 100 p&t 
cent successful takes. In their espenence heterologous immune 
serum proved just as unsuccessful as the homologous sorum. 



EFFECT OF BLOOD UPON TRANSPLANTABLE TUMORS 27 

Haaland (9) treated Berlin mice with the serum of immune 
Hamburg mice. On subsequent inoculation of the Berlin mice 
with a sarcoma to which they were normally susceptible the 
tumors grew just as well as they did in the untreated controls. 
He then repeated the experiments, employing the serum from 
immune Danish mice; but here again there was no difference 
between the growth of the tumors in treated and in untreated 
animals. 

Russell (10) and Bashford, Murray, and Cramer (11) also 
failed to obtain positive resiilts with the blood from immune 
animals. 

Sisto (12) repeated the experiments of von Dungem (13) and 
others in attempting to influence timior growth by the injection 
of splenotoxic, orchidotoxic, and hepatotoxic sera from rabbits 
into which emulsions of the fresh respective rabbit organ had 
been injected. His published charts show practically no effect 
of the injections one way.or the other. 

It would not be amiss in a paper of this character to make 
mention of a closely allied but not quite similar method of treat- 
ment that has been tried in man. In 1910 Hodenpyl (14) 
published a preliminary report of the treatment of carcinoma 
by the subcutaneous and intravenous injection of ascitic fluid 
from a patient who, he believed, had undergone a q>ontaneous 
recovery from carcinoma. All of his patients showed subjective 
improvement. It is, however, well known to those who have 
had experience id the treatment of inoperable carcinoma that 
every new form of therapy along this line is followed by a 
temporary subjective improvement, i.e., relief of pain, diminution 
of weakness, improvement in appetite, etc., only to be succeeded 
by the inevitably fatal termination. Hodenpyl stated that the 
tumors diminished in size and that some disappeared. No ciu^es 
were obtained, however, and the donor of the ascitic fluid also 
died later from cancer. 

Ill and Miningham (15) repeated Hodenpyl's treatment in a 
series of twenty-seven cases in which ascitic fluid was injected 
subcutaneously. The fluid was obtained from a patient who 
was apparently recovering from carcinoma of the liver; but 



28 ismoR KRoss 

autopsy later on revealed widespread carcinoma of the liver, 
ovaries, and intestines. The authors noted marked subjective 
improvement, although they too failed to produce a single cure 
with this treatment. 

In the experiments about to be reported, the procedure was 
as follows: The rats used were those known in the laboratory 
as the Marshall and the August breeds. The Marshall rats are 
susceptible to the Flexner rat carcinoma and refractory to the 
Jensen rat sarcoma. The August rats have the reverse char- 
acteristics in respect to these two tumors. To make doubly 
sure that the animals used were immune, a preliminary inocu- 
lation of the Flexner rat carcinoma (FRC) was made in the 
August, and of the Jensen rat sarcoma (JRS) in the Marshall 
rats. Only those animals were used which did not develop 
tumors or in which there was a small growth with subsequent 
complete retrogression. These animals constituted the stock 
from which the immime blood was taken. 

One group of Marshall rats was treated by intraperitoneal 
injections of blood from the August rats which had showed a 
natural immunity against the Flexner rat carcinoma. A second 
group of Marshall rats received an intraperitoneal injection of 
blood from normal Marshall rats. A third group received no 
preliminary injection of blood. All three groups were then 
inoculated with grafts of 0.002 gram of the FRC in the right 
axilla. Exactly the same plan was followed with the August 
rats, the immune Marshall rats being used as the source for the 
immune blood, and the JRS being inoculated. The amount of 
blood injected varied from 2 to 3 cc. depending upon the size 
of the animal, the idea being to take as much blood as possible 
without killing the rat. The tumor inoculation was made 
sjmchronously with the blood injection or within a period of 
forty-eight hours after it. 

In accordance with the theory of the workers cited, there should 
have been a complete absence of growth or at least a noticeable 
retardation of growth in those animals that had been treated 
with blood taken from the immune animals. Examination of the 
records, however, showed that the rats treated with the immune 



EFFECT OF BLOOD UPON TRANSPLANTABLE TUMORS 29 

« 

blood not only did not show any absence or retardation of growth 
of the tumor, but actually showed tumors that were in most 
instances larger, and developed earlier than those in the un- 
treated animals. Even the animals treated with normal blood 
showed this phenomenon though in a lesser degree. Gay (16) 
found that injection of blood from insusceptible or refractory 
animals lead to an increase in the number of takes of carcinoma 
hnplantation. 

Another series of experiments was made in order to determine 
whether the blood or other proteins injected act as a food. Three 
sets of rats, 36 in each group, were employed. One group 
received 1 cc. of blood (18 rats subcutaneously, and 18 intra- 
peritoneally) every three days over a period of six weeks. A 
second group received 1 cc. of egg-white (18 subcutaneously, 
and 18 intraperitoneally) every three days over a similar period. 
A third group was kept as controls. All these animals were 
inoculated with the JRS, and the tumors were charted every week 
for eight weeks. No appreciable difference was noted in the 
growth of the tumors in the three groups. Hence, the possi- 
bility that the injections acted as nutriment could be disregarded. 

CONCLUSIONS 

It is apparent from these experiments that : 

1. If immune bodies do exist in animals that are refractory 
to tumor growth, they are not resident in the circulating blood. 

2. The transfusion of blood, if it has any influence, accelerates 
the development of a tumor, as regards both time and intensity 
of growth; the inadvisability of transfusing human cancer cases 
is therefore evident. 

3. The results obtained by those investigators who report 
successful cures after injections of blood or serum must be 
explained by the assumption that they were dealing either with 
infectious granulomata. or with tumors that disappeared 
spontaneously. 

In order to reduce the effects of extraneous factors as much as 
possible, the tests were conducted in eight successive series, each 



30 



I6ID0R KBOSS 



set containing five test animals and ten controls. The test 
animals were those treated with preliminary injections of ''im- 
mune" blood, the controls those that received injections of 
normal blood or nothing at all. 



REFERENCES 



(1 
(2 
(3 
(4 
(5 
(6 
(7 
(8 
(9 
(10 

(11 

(12 
(13 
(14 

(15 
(16 



Jensen: Centralbl. f. BakterioL, 1 Abt., Orig., 1903, xxxiv, 30. 

Jensen: Ztschr. f. Krebsf orach., 1009, vii, 281. 

VON Lbtden and Bluicbnthal: Deutsch. med. Wchnschr., 1902, zrviii, 637. 

Clowes: Bull. John& Hopkins Hosp., 1905, xvi, 130. 

Clowes and Babslack: Med. News, 1905, Ixxzvii, 909. 

Cbilb and Bbbbb: J. M. Res., 1908, xiii, 385. 

v. Dunqbbn: Ztsohr. f. Immunit&tsfonch., 1910, v, 695. 

HXndel and Stbffbnhagbn: Arb. a. d. k. Gsndhtsamte., 1911, xxzyi, 490. 

Haaland: Berl. klin. Wchnsohr., 1907, zUt, 717. 

Russell: Third Sci. Report, Imperial Cancer Research Fund, London, 

1908, p. 341. 
Bashfobd, Mubbat, and Cbambb: Third Sci. Report, Imperial Cancer 

Research Fund, London, 1908, p. 315. 
Sisto: Tumori, 1918, iv, 32. 
V. Dvnobbn: Die AntikOrper, Jena, 1903, p. 30. 
Hodbnptl: Med. Rec, 1910, Ixxyii, 359. 
Ill and Miningham: J. A. M. A., 1912, lis, 497. 
Gat: J. M. Res., 1919, zv, 186. 



INOCULATION OF SARCOMATOUS TUMORS INTO 

NEGRO FOWLS,. WITH SPECIAL REFERENCE TO 

THE SIGNIFICANCE OF CHROMATOPHORES 

YUTAKA KON and TAMOTSU FUJII 

Tokio 

Received for publication December 22, 1920 

A kind of domestic fowl, called in western countries ''silky 
fowl" or "negro fowl," has its origin in the Orient. Their most 
striking characteristics are, first, the extra toes on the inner 
side of the foot, showing a complete or incomplete adherence of 
first and second of them, and second, marked pigmentation of 
every part of the body. Although the feathers of genuine negro 
fowls are pure white or pure black, the skin itself always appears 
extremely bluish brown in color and contains a number of chro- 
matophores in the cutaneous and subcutaneous connective tissue. 
The periosteum and endostetmi, dura and pia mater, perivascular 
connective tissue and serosa also show the brown coloration, 
diffuse and spotted. Of the organs the testicles and ovaries 
are most strongly tinted; the lungs show spotted pigmentation 
as in anthracosis. The parenchymatous organs, namely, the 
liver, kidney, and spleen, are usually free from abnormal pig- 
mentation. The coloration of bone is so remarkable that we 
suspected at first its identity with animal ochronosis. 

Microscopically we ascertain that the brown coloration every- 
where in the body is due to the existence of a large number of 
chromatophores in the interstitial tissue, especially along the 
course of the blood-vessels. The chromatophores assiune a 
form quite similar to those in normal fowl, having a nmnber of 
irregular long and slender processes and fine yeUowish brown 
pigment granules with an entire similarity to melanin in micro- 
scopical appearance. Pigment granules in the cells which are in 

31 



32 YTJTAKA KON AND TAMOTSU FUJII 

resting stage are very fine and equal in size. When stimulated 
by experimental means they become irregular in size, and the 
cells lose their processes, taking a round shape like any other 
plastic cells. 

The brown coloration of the bone is caused by the pigment 
cells in the periosteum. The endosteuim of the wing bone, which 
covers the inner surface of empty medullary cavity, has also a 
nimiber of chromatophores. The matrix of the bone is free 
from any kind of pigment, in contradistinction to the condition 
foimd in animal ochronosis, while the bone-corpuscles contain 
genuine melanin granules in the protoplasm, which cause a slight 
brown coloration of the bone tissue. This is especially evident 
in internal and external layers, near the peri- and endosteum. 
The bone marrow also contains chromatophores in greater or 
less number. 

The cartilage shows generally no marked coloration. The 
cartilage cells, however, sometimes contain microscopically a 
number of pigment granules. The dura mater is deep brown 
like the periosteum. The pia mater shows grajdsh brown spots 
due to pigment cells. The fascia is more or less brown in color, 
owing to a number of chromatophores which are sometimes 
to be seen in intermuscular connective tissue. The periostemn, 
pleiura, and sjoiovial membrane also show spotted pigmentation. 
The pigmentation of the testicles and ovaries is caused by the 
existence of chromatophores in the interstitial tissue, while the 
tubules and the follicles have no pigment particles. A few 
pigment cells are found in the capsules of the spleen, liver^ 
kidney, and pancreas, and in the perivascular connective tissue. 
The parenchyma of the organs and the stellate cells in the liver 
contain no pigment granules. 

The principal subject of our investigation with this kind of 
fowl was to determine whether or not the chromatophores take' 
part in the neoplastic proliferation brought about by the experi- 
mental inoculation of transplantable tumors. Such participa- 
tion, if found to occur, would be undoubtedly reliable evidence 
of the secondary acquirement of the neoplastic properties by 
normal tissue. Another object was to ascertain the origin of 



INOCULATION OF SARCOMATOUS TUMOHS " 33 

melanotic pigment in bone and cartilage cells in negro fowls, 
that is, whether it is foimed in osteoblastic and chondroblastic 
cells themselves or is given to them by adjacent chromatophores. 
Moreover, we expected this material to be most suitable for the 
purpose of studying the relation between the pigment in epi- 
thelial cells and chromatophores, which has been an interesting 
subject of discussion among authors. 



Pio. I. Beginning myxoBarcomatoua growth in the endosteum of the wing 
bone of a negro fowl, showing a number of chromatophores in the perioateum and 
endoeteum. The chromatophorea in the endosteum are deranged and replaced 
by the new grown tiaeue. Bone corpuBclea contain melanin graaulea, especially 
evident near the periosteum. 

I. THE INOCULATION OF TRANSPLANTABLE TUMOHS INTO NEGRO 
FOWLS 

Two different strains of transplantable tumors (chondro- 
sarcoma and myxosarcoma) were chosen for this purpose. Fifty 
negro fowls were inoculated, 15 of them with chondrosarcoma 
and 30 with myxosarcoma. We usually utilized for inoculation 
the empty medullary cavity of the wing bones, because the 



34 YUTAKA KON AND TAMOTSU FUJII 

endosteum is rich in ehromatophores and if any proliferation 
should take place it could be easily followed. The wing bone 
was exposed, perforated by means of a borer, and inoculated 
with the tumor. In the case of chondrosarcoma a small piece 
of tissue was used; in that of myxosarcoma, except in a few 
instances, 2 to 4 per cent watery filtrate of the tumor extract. 
The tumors were easily transplantable in the medullary cavity 
of the wing bone. Positive transplantations were obtained in 



Fio. 2. Advanced myxoa&rcomatous tumor of a negro fowl, showiog a num- 
ber of melanotic tumor cells. 

9 out of 15 inoculations with chondrosarcoma and nearly in all 
with myxosarcoma. 

The development of chondrosarcoma results in a very great 
reactive formation of callus which consists of connective, carti- 
lage, osteoid, and bone tissues. The neoplastic tissue, however, 
differs from the reactive callus tissue which is accompanied by 
a number of ehromatophores. The chondromatous cells do not 
contain any particles of pigment. Briefly, in the series of inocu- 
lations of chondroma no neoplastic change of ordinary ehro- 
matophores was seen. On the contrary, the myxomatous tumor 



INOCULATION OF SARCOMATOUS TUMORS 35 

fills up the medullary cavity of the wing bone, accompanied by 
some reactive callus tissue. In our cases metastasis was fre- 
quently to be seen, especially multiple, into lungs and medi- 
astinum, exceptionally into the skin, peritoneum, pericardium, 
and liver. The neoplastic tissue in the medullary cavities of 
wing bones is somewhat gray in color. Some of the metastatic 
tumors also show a grayish coloration. 



FiQ. 3. The aame epecimen in higher enlargement. 
Microscopically the neoplastic proliferation of endo8t«iun in 
cases of inoculation with filtrate of tumor extract seemed to 
begin in the connective tissue in the layer close to the bone 
surface. The chromatophpres in the endosteum are deranged 
and replaced by the new grown tissue. In this stage they did 
not show any sign of proliferation, having the usual form with 
a number of processes. Sooner or later, however, they began 
to multiply and produced numerous round cells with coarse 
pigment granules of irregular size. They were prolonged and 
transformed into long spindle cells which assumed an appear- 
ance quite similar to genuine tumor cells, not only in their form, 
but in their arrangement in the tumor tissue. Pigment granules 



36 YUTAKA KON AND TAMOTSU FUJI! 

are generally found close to the nucleus. Most of them, there- 
fore, are to be considered as true neoplastic elements. The 
pigment may not be a general metabolic product of negro fowl, 
for the granules are found most abundantly near periosteum and 
endosteum, where chromatophores exist in large number. This 
finding confirms that of Rous, Fujinanii, Hayashi and others, 
that the normal tissue may acquire neoplastic nature by the 



Fio. 4. Advanced myxoaarcomatous tumor of a negro (owl with an enormous 
proliferation of melanotic cells. 

inoculation of the tumor tissue or its extract. The formation 
of a pure melanotic tumor is not to be expected, for the pigment 
cells are always accompanied by a large number of the usual 
connective tissue cells, which are more readily proliferative 
than the former. In further generations of inoculation of tumor 
tissue with melanotic elements into the normal fowls the pig- 
mented tumor cells gradually disappear and the neoplastic 
proliferation of neighboring tissue cells overcomes the trans- 
planted pigmented tumor cells. 



INOCULATION OF SARCOMATOUS TUMORS 37 

II. THE RELATION BETynEEN THE MELANOTIC PIGMENT IN EPI- 
THELIAL CELLS AND CHROMATOPHORES 

With regard to the origin* of • the melanotic pigment in chro- 
matophores and in epitheliimi a difference of opinion exists. 
Some authors maintain that the pigment originates in chro- 
matophores and then goes over into epithelium, while others 
assert that it is primarily produced in epithelial cells. At the 
present time a majority of authors seem to consider that mela- 
notic pigment is formed independently both in epithelium and 
in chromatophores. 

The negro fowl offers a most suitable characteristic for the 
purpose of solving this problem. As mentioned, there are two 
kinds of negro fowl, white feathered and black feathered. Mela- 
nin granules are not to be found in the epithelium of the white 
feathered, while the black feathered variety has a number of 
pigment granules in the epithelial cells of the skin, more abun- 
dantly in the exposed parts such as the legs or the crest. In 
other parts of the body, whei*e stratum comeum is thick, the 
melanin content of epitheUum is generally Ught. The chro- 
matophores in cutaneous and subcutaneus connective tissues 
differ in numbers among both kinds of negro fowls. In this point 
the \j'hite feathered negro fowl is to be compared with the ape 
{Cercopithecus mpna), which has white hair and no pigment in 
the skin, but a large number of chromatophores in the cutaneous 
connective tissue. On the other hand the embryonal skin of 
the normal black feathered fowl shows the pigment only in the 
epithelial layer in the early stage. We recognized, therefore, 
that melanin pigment may be formed in both elements inde- 
pendently. The migration of the pigment granules from one to 
another seems to be disproved in this case, although the general 
idea that the pigment of the eye is formed in the retina and 
secondarily given to the .cells of the choroid, may not be 
impossible. 



38 YUTAKA KON AND TAMOTSU WJll 

III. THE ORIGIN OF MELANOTIC PIGMENT IN BONE CORPXJSCMBS 

AND IN CARTILAGE CELLS 

It is rather important to determine whether the above men- 
tioned pigment in bone and cartilage cells is formed in the cells 
themselves or is given from the adjacent chromatophores. For 
this purpose I have intentionally broken the ulna of negro fowls 
in one or both sides and examined the callus histologically in the 
course of one to forty-three days after fracture. Twelve negro 
fowls were used for this experiment. During healing of fractured 
bones the callus formation takes place at first close to the bone 
tissue, stripping off the layer of chromatophores. Periosteal 
chromatophores are arranged close to each other in the earlier 
stage. With the advance of callus formation they are separated 
by new grown tissue cells. Actual proliferation of chromato- 
phores takes place in some degree, forming round shaped cells 
with irregular melanin granules. At the same time melanin 
granules are to be found in some typical osteoblasts and chondro- 
blasts in the callus, but far less in number than in new formed 
chromatophores. Pigment granules in the plastic cells are 
generally coarse in size, but they do not show any microchemical 
difference from those in chromatophores. The amount of pig- 
mented plastic cells is generally proportional to the number of 
chromatophores in the body. There is no actual sign of trans- 
ference of pigment from chromatophores to plastic cells. The 
metaplastic change of chromatophores to fixed tissue cells is 
not recognizable, for the latter are to be acknowledged undoubt- 
edly as a sort of wandering cell. We conclude, therefore, that 
genuine osteoblasts and chondroblasts themselves can make 
melanin into protoplasm, and that the melanin formation may 
occur in many kinds of tissue cells, if there are some melanogenetic 
compounds from which melanin could be derived, and fermenta- 
tive (perhaps oxidative) elements whigh act on them. 

SUMMARY 

1. Oriental negro fowls have a large number of chromatophores 
in all parts of the body, especially in the connective tissue, dura 



INOCULATION OF SARCOMATOUS TUMORS 39 

• 

and pia mater, periosteum and endosteum, serous membrane, 
interstitial tissue of genital glands, and in the lungs. 

2. The bone corpuscles and occasionally the cartilage cells 
also contain melanotic pigment in their protoplasm. 

3. Myxosarcomatous tumors of negro fowls grown by inocu- 
lation of tissue or its extract assume a melanotic nature, probably 
due to the neoplastic proliferation of chromatophores. This 
is actual evidence for the acquisition by normal tissue of neo- 
plastic nature. 

4. The melanotic pigment in the epithelium and connective 
tissue is formed in both elements independently of each other. 
The migration of pigment granules from one to another is not 
recognizable. 

5. Osteoblasts and chondroblasts in callus also form melanotic 
piionent in their protoplasma. 



INFLUENCE OF THE LYMPHOCYTE ON THE PERI- 
TONEAL IMPLANTATION OF SARCOMA IN MICE 

ELLIS KELLERT 
From the Bender. Hygienic Laboratory, Albany, New York 

Received for publication January 22, 1921 

No phase of tumor investigation in recent years has aroused 
greater interest than that of the relation of the lymphocyte to 
natural and induced resistance to cancer. Following the obser- 
vation by Da Fano (1) that lymphoid and plasma cells are 
constantly associated with degenerating tumors and that the 
lymphocyte might be the agent which distributes immunity 
throughout the organism, many articles have been written 
tending to prove an antagonistic action between the lymphocyte 
and tumor growth. Burgess (2) noted that the retrogressing 
implant of cancer ui mice is invariably surrounded by large 
niunbers of lymphocytes. Similar observations were made by 
V. Dungem and Coca (3). Apolant (4) found increased difficulty 
in immxmizing splenectomized animals to tumor implants. 
Tyzzer (5) working with a Japanese waltzing-mouse tumor found 
that implants grew vigorously for a time in non-susceptible 
mice, but were destroyed soon after by an inflammatory reaction 
excited only by Uving tumor; the cell infiltration disappearing 
as the tumor became necrotic. In previously immunized non- 
susceptible mice the process of destruction was the same but 
began much earlier. Braunstein (6) concluded that the spleen 
possesses highly developed powers of resistance against tumor 
growth, while splenectomized animals are more liable than normal 
animals to inroads of malignant growth. Rohdenbui^ and 
Johnston (7) made parallel observations upon thymus, pancreas, 
spleen, hypophysis, and testis, showing that after extirpation of 
the th}rroid, thymus, or testes, the animal, had a lessened resist- 
ance to carcinoma. 

41 



42 ELLIS KBLLERT 

Baeslack (8) studied the blood of inoculated animals and 
reported a decrease in the nxunbers of circulating lymphocytes 
in mice with growing tumors. Murphy (9), working with tissue 
cultures, noted that resistance to heteroplastic tissue grafts 
apparently depends upon the activity of the lymphocjrte and 
obtained similar results with experimental cancer in mice. 
Murphy and Morton (10) in a series of experiments with trans- 
planted and spontaneous tumors in mice noted that slight expo- 
sure to Roentgen rays, inducing a lymphocytosis, brought about 
a more highly resistant state to implanted tumor; whereas larger ; 

doses of the x-ray, inducing a leucopenia, caused a decreased ■ 

resistance to inoculation with tumor, and that animals naturally j 

immune were rendered susceptible by this method. They con- 
cluded that destruction of the l3rmphocytes by the x-ray causes 
a loss of the natural or induced resistance to the growth of inocu- 
lated cancer. 

The apparently conclusive results of the investigators referred 
to above have not been without contradiction. Wade (11) 
noted that with the establishment and growth of an inoculable 
tumor, there is a steady increase in the percentage of lympho- 
cytes in the circulating blood. Rous and Murphy (12), experi- 
menting with sarcoma of the fowl, decided that no fimdamental 
importance can be attributed to the protective function of 
l3rmphocytes in mammals and birds alike although they did 
state that the '^lymphocyte has an association with the process 
of resistance in the fowl more marked than that seen Inmammals.'' 

Serafini (13) observed that ligation of the splenic vessels in 
rats and the injection of splenic tissue does not promote the 
growth of implanted tumor. Jones and Rous (14) noted that 
mice inoculated eight to ten days after splenectomy have the 
same resistance as normal animals; and that mice inoculated 
three and a haH weeks after splenectomy live longer than normal 
animals. Bullock (15) and Stevenson (16) could not find that 
the presence of spleen inhibited the growth of tumor implanted 
into the chick embryo. Sittenfield (17), having utilized various 
means of inducing a hyperlymphocsrtosis, stated that ^'neither 
increase nor reduction of the lymphoid elements in the blood 



PERITONEAL IMPLANTATION OF SARCOMA 43 

had any influence upon either resistance or susceptibility to 
tumor growth." Stevenson (18), again, found in 1917 that 
tumors show no inhibition of growth when grown on the allantois 
of the chick embryo for ten days in the presence of chicken 
spleen. Bullock and Rohdenburg (19) (20) also reported that 
removal of the spleen does not favor the growth of heteroplastic 
tumor grafts. More recently Prime (21) found that, following 
lymphocytic reduction by exposure to the x-ray, no appreciable 
decrease in immxmity to tumor inoculation occurs. 

As a result of these conflicting and widely divergent residts 
we find that the rdle of the lymphocyte in immunity to cancer 
is still imdetermined. The varying conclusions arrived at are 
due to the methods employed and to the instability of the 
leucocyte count in the lower mammals, for blood-cell coimts in 
mice may be extremely variable. In attempting to devise a 
new means of approaching the problem, it occurred to the author 
that intraperitoneal inoculation of a tumor, preceded and fol- 
lowed by study of the cellular elements of the peritoneum, might 
yield information regarding the part played by the lymphocyte 
in cancer immunity. 

The normal peritoneimi' of the white mouse contains numerous 
cellular elements, among which are found in predominating 
numbers the large mononuclear phagocjrtic cells designated as 
macrophages and small mononuclear cells having the morphology 
and staining reactions of lymphocytes. In addition there are 
large cells whose cytoplasm is filled with coarse basophile granules, 
resembling the mast cells of the blood, a few polynuclear leuco- 
cytes, and erythrocytes rarely. The macrophages, as nearly 
as can be determined by direct counts, are found to vary from 
10,000 to 20,000 per cubic millimeter,, whereas the lymphocytes 
are far more numerous, usually about 115,000 per cubic milli- 
meter. In a few mice the macrophages predominate, but in 
the majority the lymphoid cells are more numerous. In differ- 
ential counts made by obtaining peritoneal fluid with capillary 
glass pipettes, the average percentages found in a series of 24 
mice were as follows : 



44 ELLIS KELLERT 

percent 

Macrophages 34.3 

Large mononuclears 7.0 

Lymphocytes 66.0 

Polymorphonuclear neutrophfles 3.5 

Basophi es 0.2 

100 

The large mononuclears in the above classification are much 
larger than the lymphoid cells, but do not possess the abundant 
foamy cytoplasm of the macrophage. The designation of the 
small mononuclear cells as lymphocytes may be questioned. 
Evans (22) apparently refers to certain small mononuclears in 
the peritoneum as young macrophages, asserting that they will 
take up granules of vital stains, a characteristic lacking in true 
lymphocytes. This may apply to the cells referred to above as 
large mononuclears. Morphologically we find the small cells 
to be identical with the lymphoid cells of the blood. We do 
not find that they possess phagocytic properties for bacteria, 
inert particles, or vital stains; and finally, they decrease markedly 
under x-ray exposures. 

It is very evident, then, that the peritoneum of the mouse 
contains large niunbers of cells having all the characters of 
lymphocytes, and it is the purpose of the present investigation 
to show their influence on intraperitoneal inoculations of a 
mouse tumor obtained from the Crocker Research Fund, where 
it is designated as no. 180, and is regarded as a sarcoma. The 
tumor grows readily, and in our hands has not failed in more than 
5 per cent of subcutaneous and 2 per cent of intraperitoneal 
inoculations. Inoculations were made by means of a small 
trocar needle. 

The white mice selected were young adult animals; an equal 
number of each sex. Smears from the peritoneiun were made 
before inoculation and repeated after inoculation till the "taps" 
became bloody and unfit for differential counting. Considerable 
variation in the percentage of Ijonphocytes occurred in the 
inoculated mice, and also some variation in the same mouse at 
different times. The lowest differential coimt recorded showed 
25 per cent lymphocytes, the highest 80 per cent of lympho- 



PEMTONBAL IMPLANTATION OP SARCOMA 



45 



cjrtes, and the average in uninocnlated mice 66 per cent lympho- 
cytes. After inoculation^ when the tumors were well developed 
and readily palpable through the skin, there was found an 
average of 54 per cent of Ijrmphocytes. Twenty-two mice were 
used in this experiment. About half the animals showed a rise 
and the other half a fall in the percentage of lymphocytes fol- 
lowing tumor inoculation. Table 1 illustrates the results 





TABLE 1 




• 


PXBCEMTAOK OF LTllFBOOTrBS 




IIOUBB 




TXntOR 




Before inoculation 


After inoculation 




18 


60 


75 


Large 


19 


50 


74 


Large 


20 


30 


56 


Large 


21 


48 


45 


Large 


24 


80 


75 


Large 


27 


66 


75 


Large 


30 


60 


64 


Large 


32 


64 


45 


Large 


33 


65 


85 


Large 


34 


44 


56 


None 


35 


61 


32 


Large 


36 


54 


40 


Large 


37 


50 


29 


Large 


47 


56 


53 


Large 


67 


48 


50 


Large 


68 


66 


70 


Large 


69 


44 


58 


Large 


70 


49 


31 


Large 


71 


60 


50 


Large 


74 


80 


70 


Large 


76 


46 


29 


Large 


77 


40 


42 


Large 



Of 22 mice, 11 show an increase in lymphooytesi and 11 a decrease. 

obtained before and after inoculation. f*ollowing the inocula* 
tion, numerous polymorphonuclear leucocytes appeared in the 
peritoneum but not in sufficiently large numbers to influence 
visibly the percentages of the other cellular elements. 

Thus, as indicated in the table, of 22 mice inoculated with 
tumor all with one exception presented in a short time (ten to 
twenty days) rapidly growing tumor nodules. In making the 



46 ELLIS KELLERT 

punctures with capillary glass tubing blood was frequently 
obtained and when abundant the slides were discarded. The 
counts recorded above were made on relatively blood-free 
preparations. Since many mice showed an increase and others 
a decrease in the numbers of lymphocytes, and since the tumors 
were large and growing rapidly, it is apparent that no marked 
adverse influence can be attributed to the presence of the lympho- 
cytes. The most important point brought out by the present 
investigation is the fact that small fragments of tumpr tissue 
placed in a fluid rich in lymphoid cells showed no inhibition 
of growth. 

CONCLUSIONS 

1. The normal peritoneum of the white mouse is rich in cellular 
elements — 115,000 to the cubic millimeter — of which about 55 
per cent are lymphocytes. 

2. Mouse tumor implanted into the peritoneum of these mice 
grows rapidly. 

3. No marked change in the cellular content of the peritoneum' 
of such mice results from the inoculation of mouse tumor. 

4. No direct antagonistic action is foimd to exist between the 
lymphocyte and tumor implants in mice. 

REFERENCES 

(1) Da Fano: Ztschr. f. Immunit&tsf orach., Orig., 1910, v, 1. 

(2) BuROESs: Jour. Med. Res., 1909, N.S., xvi, 576. 

(3) v. DuNOERN AND Coca: Ztschr. f. Immunitfttsforsch., Orig., 1909, ii, 391. 

(4) Apolant: Ztschr. f. Immunit&tsforsch., Orig., 1913, xvii, 219. 
(6) Ttzzer: Jour. Med. Res., 1916, N. 8., xxvii, 201. 

(6) Braitnstein: Quoted by Oser and Pribram, Amer. Jour. Med. Sci., 1913, 

cxlvi, 310. 

(7) RoHDENBURG AND Johnston: Quoted by Oser and Pribram, Amer. Jour. 

Med. Sci., 1913, cxlvi, 310. 

(8) Baeslack: Ztschr. f. Immunitfttsforsch., Orig., 1914, xx, 421. 

(9) Murphy: Jour. Exper. Med., 1913, xvii, 482. 

(10) Murphy and Morton: Jour. Exper. Med., 1915, xxii, 204, 800. 

(11) Wade: Jour. Path, and Bacteriol., 1908, xii, 384. 

(12) Rous AND Murphy: Jour. Exper. Med., 1912, xv, 270. 

(13) Sbrawni: Tumori, 1917, v, 229. 

(14) Jones and Rous: Jour. Exper. Med., 1914, xx, 404. 



PERITONEAL IMPLANTATION OF SARCOMA 



47 



(15) Bullock: Lancet, London, 1915, i, 701. 

(16) Stevbnbon: Jour. Cancer Res., 1917, ii, 245. 

(17) Sittenfield: Jour. Cancer Res., 1917, ii, 151. 

(18) Stbyenbon: Jour. Cancer Res., 1917, ii, 449. 

(19) Bullock and Rohoenbxtbg: Jour. Cancer Res., 1917, ii, 465. 

(20) Bullock and Rohdenburg: Jour. Cancer Res., 1918, iii, 31. 

(21) Primb: Jour. Cancer Res., 1920, y, 106. 

(22) Evans: Amer. Jour. Physiol., 1915, xxxvii, 243. 



PLATE 1 



Fios. 1 AND 2. Examples of rapidly growing peritoneal tumors following tumor 
inoculation into mice with numerous lymphocytes in the peritoneum. 



48 



PERITONEAL IMPLANTATION OF 3ARCOUA 






• V 



OCT 23 19 2 



LlBRfi^i^^ 



PLATE 2 

Figs. 3 and 4. Peritoneal smears from normal mice. Obj. 16 mm. Oc. 5. 
Note the large numbers of lymphocytes. The larger pale-staining cells are macro- 
phages. A few mast cells are present. 



50 



PERITONEAL IMPLANTATION OF 8.\RC0MA 



PLATE 3 

Fio. 5. Peritoneal smear showing large numbers of macrophages. Tumor 
growth rapid. Obj. 4 mm. Oc. 6. 

Fig. 6. Peritoneal smear from mouse which showed 75 per cent of lympho- 
cytes in presence of well-developed peritoneal tumor. Obj. 4 mm. Oc. 5. 



52 



PERITONEAL IMPLANTATION OF SARCOMA PLATE 3 

ELLIL KELLERT 



• 









• "^ • • ^^ ••• •« •• 


















Fig. 5 Fig. 6 



53 



PLATE 4 

Fig. 7. Peritoneal smear showing 75 per cent lymphocytes. Tumor large and 
rapidly growing. Obj. 4 mm. Oc. 5. 

Fig. 8. Peritoneal smear showing 70 per cent lymphocytes. Tumor large. 
Obj. 4 mm. Oc. 5. 



54 



PERITONEAL IMPLANTATION OF SARCOMA 



• » *• ' • * • 

•*•% • • • • 

.. . *• * » • 

■•*■•. 



PRIMARY SPONTANEOUS SQUAMOUS CELL 

CARCINOMAS IN MICE 

STUDIES ON THE INCmENCE AND INHERITABILITY OF 

SPONTANEOUS TUMORS IN MICE 

Fifteenth Communication 

MAUD SLYE> HARRIET F. HOLMES, and H. GIDEON WELLS 

From the Oiho S. A. Sprague Memorial ImtitxtU and the Department of Pathology 

of the University of Chicago 

Received for publication April 4, 1021 

Throughout most of the animal kingdom primary carcinoma 
of the skin is met with but rarely, in contrast with its great 
frequency in man, and also with the relative frequency of other 
cutaneous and subcutaneous tumors in animals. The existing 
literature on spontaneous tumors in animals indicates that no 
species except dogs, horses, and fowls, show skin carcinoma 
except as a rare condition, and this is of particular significance 
in view of the fact that skin carcinomas are more certain to be 
recognized than those in any other location and hence their 
relative infrequency is even more marked than any statistical 
evidence woidd indicate. Presumably the improtected con- 
dition of the human skin accounts for its susceptibility to cancer, 
as the hairy scalp rarely shows primary carcinoma, and the 
pigmented skin of the negro also seems insusceptible. 

The classical compilation of Sticker (1), which contains records 
of many tumors not fully differentiated according to modem 
criteria, gives the following relative frequency in this material 
of skin tumors in different species: 



57 



58 



MAUD SLTE, H. F. HOLMES AND H. G. WELLS 



Total carcinomas 

Skin 

Eye 

Lip 

Anus 

Penis 

Vulva 



Hone 



332 
22 
14 
4 
8 
52 
11 



Cattle 



78 
2 
2 
1 

2 
1 



Dog 



766 

166 

8 

89 
16 



Gat 



21 
6 



1 
1 



Swine 



12 
2 



The extensive compilation by Teutschlaender (2) of the distri- 
bution of tumors in animals shows that ca^es of squamous cell 
carcinoma have been described in but few species, according to 
his records, namely : cat, dog, mouse, rat, elephant, beef, sheep, 
pig, horse, mule, cockatoo, toucan, chicken, carp, and gold fish. 
However, in another table he mentions also a case in a stag. 

Apparently birds develop cutaneous carcinoma somewhat 
more often than most other animals, but even here there are 
but few recorded. In their summary of the literature on tumors 
in birds, Joest and Emesti (3) found but 4 of 37 tumors described 
in birds to be skin carcinomas, those being two found on the legs 
of chickens by Wernicke; one on the wing of a "papapeis" — 
reported by Guerrini, and one on the wing of a toucan reported 
by Herbert Fox (4), this last case exhibiting also a metastasis 
in the lung. In none of their own 50 cases of tumors in birds 
was a skin carcinoma described. This compilation, as well as 
all others that we have observed, omits the strikmg case of 
squamous carcinoma of the skin of the foot of a hen described 
by Boynton (5), in which transplantations were made without 
success. More recently a similar tumor arising on the thumb 
claw of a lark has been reported by Urra (6). Two cases of 
squamous cell carcinoma of the mouth in fowls have been 
described by Pick and Koch. The experience of Teutschlaender 
(7) is unique, for he reports that of 54 carcinomas in fowls 
observed in the cancer laboratory at Heidelberg, 28 were skin 
epitheliomas, of which 18 were on the foot, attributed to the 
irritation of parasitic infection of the epidermis {Cnemidocop^ 
iesmUben). 



SQUAMOUS CELL CARCINOMAS IN MICE 59 

Apparently fish are subject to skin tumors of various structure, 
including carcinoma. Thus, Fiebiger (8) reports cases of epi- 
thelioma occurring on the lip of each of two "Schlei" {Tinea 
vulgaris Cuvier) coming from the same pond, a markedly malig- 
nant skin epithelioma in a carp, and mentions papillomatous 
growths in several specimens of Eletterfish {Anabus scandens). 
Plehn has also described nmnerous skin epitheliomas in cyprinoid 
fishes. 

Bashford has reported a squamous ceU carcinoma in a small 
fish, Gaaterosteur apinachia, two cases of carcinoma of the skin 
in frogs, and a carcinoma of the skin glands in a Triton. 

Among mammals the dog alone seems to be commonly sub- 
jected to skin carcinoma, if we except the carcinoma of the penis 
and prepuce occurring not infrequently in horses and oxen. It 
would seem to be not imcommon in cats according to the experi- 
ence of Roffo (9), who found 7 tumors in 307 cats examined, of 
which 4 were skin carcinomas on the head and face, with Ijrmph 
gland metastases, but none of 11 cases of tiunors in cats col- 
lected by Bashford (10) was of this tjrpe, and but 2 of 48 tumors 
in dogs, although there were several squamous cell growths in 
the mouth and pharynx. Leo Loeb (11) reported in 1903 that 
from one to three carcinomas of the inner canthus of the eye 
were observed every week in the Chicago stock yards, but in 
one year no other external carcinomas were observed except 
two of the vulva. 

Rodents seem to have skin carcinoma but rarely. The large 
series of wild rats autopsied in plague work and examined for 
tumors (McCoy, WooUey and Wherry, Beatti) have revealed 
no such growths except for an epithelioma of the vulva reported 
by Woolley and Wherry (12). It will be recalled that the early 
transplantation experiments of Hanau (13) were with a squamous 
cell carcinoma from the v\ilva of a white rat, and he states that 
there had been two previous cases of similar growths in rats in 
their laboratory. Roffo (14) has also observed a squamous cell 
carcinoma which arose in the vulva of an old white rat, which was 
transplanted through six generations without change in structure. 
In view of the great frequency of parasitic skin infection in rats. 



60 MAUD SLYE, H. F. HOLMES AND H. G. WELLS 

''rat scabies," commonly with extreme papillomatous prolifer- 
ation of the epitheliimi (15), it is strange that epithelioma has 
not been more frequently observed. Of 123 cases of spontaneous 
rat tumors reported in the literature and collected by Bullock 
and Rohdenburg (16), there were but three epitheliomas, one of 
the tongue and two of the vulva. They reported on 48 rat 
tumors observed in ,the Crocker Laboratory, one of which was 
an epithelioma of the head, which was not described further. 
Morris (17) has reported a basal cell carcinoma which arose in 
the skin of a male rat, five months old, and which died out after 
two generations of transplants. He was tmable to find other 
recorded instances of basal cell carcinomas in animals. 

Several instances of squamous cell carcinoma have been 
. described in mice, but they form an insignificant proportion of 
the many spontaneous tumors that have been observed in this 
species. Nevertheless they were looked upon as of great signifi- 
cance in the early days of experimental cancer research, since 
they established the fact that mice had other forms of tumor 
than the predominating mammary gland carcinoma. Haaland 
(18) reported a case of squamous cell carcinoma arising in the 
mouth of a mouse in Borrel's laboratory, involving the lower 
jaw, showing a typical squamous cell structure. Six mice were 
inoculated from this, but as in six months they showed no 
growths they were put back into the cage with the others. Ten 
months after the inoculations ''deux cas identiques de cette 
m^e tumeur apparurent en m^e temps dans cette cage," 
but it was not known whether these were inoculated mice or not 
and Haaland does not make clear whether these later tumors 
were in the mouth, although this is apparently what he means. 
Later he observed a third case of mouth cancer in a mouse 
recently inoculated with a Jensen timior. No growths developed 
in mice inoculated from this mouth tiunor. 

Murray (19) in 1908 described the following cases of super- 
ficial squamous cell cftrcinoma in mice: (1) A prickle cell growth 
with apparently little keratinization, arising in the skin of the 
neck, with metastasis into an adjacent lymph gland. (2) A 
carcinoma presenting, both alveolar and squamous cell areas, 



SQUAMOUS CELL CARCINOMAS IN MICE 61 

with keratinizing metastases in the lung. Presumably this tumor 
arose in the mammary gland. (3) Prickle cell growth without 
keratinization, interpreted as primary in the nipple; this mouse 
also had had a hemorrhagic adenocarcinoma of the mammary 
gland removed by operation. In discussing the mammary 
gland tumors he also describes the occurrence of areas of keratin- 
ization within tubular carcinomas, and regards them as indi- 
cations of the close association of the mammary apparatus with 
the skin from which it develops, and points out that the ampulla 
which receives the terminal portions of the mammary ducts is 
also lined by stratified squamous epithelium. ' ' Therefore, should 
the cells of the new growth have taken their origin near the 
nipple, variations in either direction are only to be expected. " 

In his later report from the London laboratories, Haaland 
(20) states that of 353 spontaneous tumors observed (not 
including those in Murray's report), 22 were squamous-celled 
carcinomas with marked keratiQization, of which 14 arose in 
l^e mammary region and 8 outside it. Of these 8, 3 were mouth 
tumors similar to those observed in Borrel's laboratory, appar- 
ently arising in the mucous membrane of the inside of the cheek, 
there being in one metastases in the lymph nodes. One seemed 
to have arisen on the surface of the head, and it penetrated the 
skull; grafts from this were successful. (The illustrations of 
these tumors show the gross relations of such tumors so well 
that we are spared the necessity of reproducing further illus- 
trations.) Three tumors were described as superficial verrucous 
growths, on the vulva, anus, and left flank. One was a typical 
squamous cell carcinoma of the skin of the chest, developing as 
a superficial ulcerating growth, ''undoubtedly from the skin 
itself, or from the mammilla." A case of mixed squamous and 
sebaceous carcinoma of the vulva is also mentioned. 

Several of the keratinizing mammary tumors differ from 
typical squamous cell carcinomas, and their cutaneous origin is 
doubtful. Four are described as resembling "molluscoid" tumors, 
characterized by long radiating cylinders of cells with a central 
keratinized zone, found in the mammary region and not giving 
successful transplants. Four tumors are described as ''typical 



62 MAUD SLYE; H. F. HOIiMES AND H. G. WELLS 

squamous-celled alveolar carcinomata in the mammary region. " 
Six others are called ''adenocancroids," these being imques- 
tionably mammary gland adenocarcinomas in which areas of 
keratinization occur. Attention is called to the occurrence of 
squamous-celled cysts in the normal mannna, apparently arising 
from mammary ducts in which the epithelium has become 
metaplastic, and which perhaps explains the occurrence of 
keratinizing carcinoma in the mammary gland. As squamous- 
celled growths have also been found rarely in the himian breast, 
as well as in cats and dogs — ^Teutschlaender (21) — ^the disease 
in mice is not without analogy, although apparently relatively 
'frequent among the mammary gland tumors of this species. 
Presumably in some stocks of mice keratinization is even more 
common in mammary gland carcinoma than the above reports 
indicate, for Woglom (30) foimd this change in 228 of 1000 
spontaneous mammary gland carcinomas examined in the Crocker 
Laboratory. 

Pick and Poll (22) reported as a ''sweat gland carcinoma'' a 
tubular growth arising near the scapula of a mouse, but the 
description published does not permit the exclusion of a mammary 
gland origm. 

Erdheim (23) described a pedunculated tiunor, which gave no 
evidence of malignancy, but exhibited a histological structure 
resembling squamous cell carcinoma. It arose from the vulva 
of a mouse, and after the outer part of the growth had been 
removed the base healed spontaneously. He was imable to 
classify this tmnor, beyond indicating that it was a stratified 
epithelial neoplasm. 

Tsutsui (24) has produced carcinoma on the skin of the back 
of mice by painting with tar, after the procedure used by Yama- 
giwa with the ears of rabbits, observmg 16 carcinomas and 1 
sarcoma in 17 nuce siuriving over 100 days. In two cases lung 
metastases were observed. 

We find no other reports of squamous cell carcinoma arising 
in the skin or mouth of mice, which indicates their relative 
infrequency, since such tumors are most obvious when they do 
occur. Thus, Tyzzer (25) in his report of 83 spontaneous 



SQUAMOUS CELL CARCINOMAS IN MICE 63 

tumors in mice includes no timiors of this type. In the 41 
primary mouse tumors described by Jobling (26) there were no 
skin epitheliomas, although there was one of the molluscoid 
keratinizing tumors such as Murray, Haaland, and Tyzzer have 
described. 

Squamous cell carcinomas outside the mammary gland and 
skin of mice are rare. Tyzzer reported that among his lung 
tumors ''in several cases" the growth was of an epidermoid 
character, and Haaland described one such tumor. In our 
previous paper on primary carcinoma of the stomach in mice 
(27), we collected four cases of squamous cell gastric carci- 
noma reported in the literature and added three more in our 
own material, as well as one case of squamous carcinoma in the 
external siuf ace of a chronically prolapsed rectum. Fiebiger 
(31) has also reported the occurrence of a few instances of 
squamous cell carcinoma produced experimentally in the stomach 
of mice by feeding cockroaches injected with Spiroptera. We 
have found no reports of squamous cell carcinomas arising in the 
esophagus or urinary tract, or in the cervix uteri, or in any other 
structure where they might be found in mice, except such as 
have been mentioned above. 

OBSERVATIONS ON SKIN TUMORS AND SQUAMOUS CELL 
CARCINOMAS IN THE 8LYE STOCK OF MICE 

In 28,000 consecutive autopsies performed on mice of this 
stock, which had been permitted to live as long as posisible 
without any experimental manipulations whatever, we have 
observed the following instances of primary neoplasms of squa- 
mous or stratified epithelial structure. 

Primary squamous cell carcinoma of the skin and mouth 70 

Primary basal cell carcinoma of the skin 15 

Primary keratinizing carcinoma of the mammary gland 56 

Primary squamous cell carcinoma of the stomach 4 

Primary keratinizing carcinoma of the lung 1 . 

Primary squamous cell carcinoma of the rectum 2 

Primary squamous cell carcinoma of the vagina 1 

Primary stratified carcinoma of the meibomian gland 2 

Primary sebaceous gland adenocarcinoma 1 

152 



64 BiAUD SLYE, H. F. HOLMES AND H. G. WELLS 

As we have not yet made a complete analysis of all the timiors 
observed in these 28,000 autopsies the proportion of stratified 
cell tmnors to total tumors cannot be stated. Roughly there 
are probably about 4000 primary spontaneous tumors of all sorts, 
so that these squamous cell tiunors constitute not far from 4 per 
cent of the total. 

CARCINOMA OP THE SKIN 

In our material are 85 malignant epithelial growths arising in 
the skin, mouth, and lips, excluding those squamous cell growths 
that seem to be derived from mammary gland tissue and strat- 
ified cell growths from other glands. Of these, 70 are of squamous 
cell type and 15 of basal cell character. Since many of the 
tiunors which arise about the lips and mouth are, when observed, 
so extensive that their exact point of origin cannot be deter- 
mined, we have grouped them together with the skin tumors. 
Roughly plassifying the site of the 70 squamous cell growths, 13 
were iqi the skin of the trunk, one on a front limb, 2 on the vulva, 
15 >ibout the lower jaw, IS on some other portion of the face, 
and 23 about the ears or neck; that is, all but 16 of the 70 were 
on the head and neck. Presumably this is to be explained by the 
much greater amount of traumatism suffered by the skin in this 
part of the body, which is especially marked in cage mice which 
are always rubbing their muzzles against the rough wire meshes 
of the cages. The tumors of the jaw and mouth in several 
instances seemed to have resulted from the irritation produced 
by broken or protruding teeth. Nearly all of the carcinomas 
arising on the tnmk were definitely located at the site of a healed 
wound, and in not a few of those of the head the same origin 
was observed ; it is probable that less obvious woimds are respon- 
sible for many if not all of the others. A chronic dermatitis 
often preceded the carcinomas of the skin of the face, presumably 
incited by traumatism. The not uncommon fungus infections 
of the skin seem to be too rapid in their course to lead to carci- 
noma formation. 



SQUAMOUS CELL CARCINOMAS IN MICE 65 

Clinically these growths are similar to corresponding forms of 
carcinoma in man (fig. 1). The age at which they occur is, on 
an average, later than with any other mouse tumors, as they 
rarely appear before late middle life, and usually only in old 
age. They arise under crusted ulcers, or develop in areas of 
hyperplasia from chronic irritation, spread slowly as crusting, 
ulcerated growths, destroying at times the eyes or other features, 
usually remain superficial, and most often cause death by chronic 
infection of the ulcerated surface, which may also lead to acute 



abscess formation. Occasionally death results from starvation 
when the cancer involves the mouth or jaw, from hemorrhage, 
invasion of the skull, or other coincident diseases. 

The microscopic diagnosis of the skin neoplasms is often 
difficult, especially with the basal cell growths, since we find all 
stages of epithelial overgrowth from simple hyperplasia to 
metastasizing carcinomas. In this series no growth is included 
as a carcinoma unless it exhibited both the gross and clinical 
features of malignancy together with distinct microscopic 
e\idence of infiltrative character. The extensive destruction of 
these growths by infective processes often obscures the micro- 
scopic findings. In several instances a growth that presented 



p THE Jaw, with Formation 



FILTRATING Adjacent Tissues 
ATI ON. No. 26238. X 60 



SQUAMOUS CELL CARCINOMAS IN MICE 67 

all the clinical features of carcinoma has been found at autopsy 
so extensively necrotized and suppurating that only examina- 
tion of numerous sections from different parts of the growth 
has established the diagnosis of carcinoma. It is quite possible 
that we have lost a few genuine carcinomas through such destruc- 
tive processes, to say nothing of cases in which the cannibal- 
istic mate has selected the neoplasm for the first course. 



Fio. 4. Squamous Cell Carcinoma of Skin or Chest Arisino in the I 
OP AN Old Wound, the Scar Tissue Bbino Abundant in this Field, 
AND THE Relation op the Cancer Growth to the Over- 
lying Skin Being Shown. No. 8212. X 45 

The microscopic features of the squamous cell carcinoma 
the skin and mouth of mice differ not at all from those that 
occur in man. Usually homification is marked (fig. 2), but, as 
in man, growths from a similar origin may show little or no 
tendency to form keratin (fig. 3). Usually the amount of 
connective tissue formation is not large, except in some cases 
where the cancer has developed in scar tissue from old wounds 
(fig. 4). Little help in diagnosis is afforded by mitotic figures, 



68 MAUD SLYE, H. F. HOLMES AND H. G. WELLS 

since these are scanty in even the most typically malignant 
growths, and commonly cannot be found at all. Presumably 
they would be much more numerous if the tumors were removed 
from the living animal. . 

The infiltrative character of growth is not usually so extensive I 

as we are accustomed to see it in man, for the mice generally !, 

succumb to infection from the ulcerated growth while it is 



Fio. 5. Squamous Cell Carcinoma of the Head and Face Invading the 

Spinal Column in the Cervical Reoion. Comprebsino the 

Spinal Cord and Inhltratino the^Mbninqes 

This tumor also invaded the skull. No. 23766. X 60 

still small. Infiltration is usually seen best in the tumors that 
arise about the jaws, for here the bone is often invaded. Bone 
infiltration was observed in 7 of the 15 tumors that arose about 
the jaw. We have had one striking case (23766) in which a 
carcinoma of the skin of the head, beginning at the base of the 
right ear, infiltrated the skull and cervical vertebral canal, 
infiltrating the meninges about the cerebellum and compressing 
the spinal cord in the cervical region (fig. 5). This case resembles 



SQUAMOUS CELL CARCINOMAS IN MICE 69 

the one described and pictured by Haaland, in which the skull 
was invaded. In two instances we have seen infiltration of the 
salivary glands, and in three cases an adjacent lymph node 
was involved apparently by direct extension. 

The absence of Ijmiphatic metastasis is a striking feature 
of these tumors when compared with corresponding growths in 
man, which is true of all forms of carcinoma in mice. In only 
two cases did we find a secondary growth by metastasis into a 
lymph node from a skin carcinoma (7950, 15232), and it was a 
very common observation that lymph nodes immediately 
adjacent to or in contact with these squamous cell growths, 
were not involved. Possibly serial sections of all our mice would 
have revealed other instances of metastasis, but the value of 
the information did not seem commensurate with the labor 
involved. In only one case did we find visceral metastasis 
(12627). This mouse had a squamous cell carcinoma arising 
just dorsal to the rectum, and a t3rpical secondary nodule about 
1 mm. in diameter in the lung; it also had a tubular carcinoma 
of the mammary gland which had produced no metastasis. The 
infrequency of metastasis in this series of spontaneous squamous 
cell carcinoma is significant when compared with Yamagiwa's 
experimental tar tumors in rabbits, since the infrequency of 
metastasis in his material has been thought by some to speak 
against their being true malignant tumors. As a matter of fact 
the metastasis incidence obtained by him, and by Tsutsui with 
experimental skin tmnors in mice, is distinctly higher than that 
observed in these spontaneous growths. 

The sex incidence is strongly in favor of the female (49 to 21) 
in this series. As most of the growths were on the head and 
neck we cannot account for this on the basis of erroneous inclusion 
of squamous carcinomas of the mammary gland with our skin 
tumors. It differs from our experience with other growths of non- 
reproductive tissues which have shown approximate equality as 
to sex incidence. Even more than with other tumors we have 
found that age is an important factor, these squamous cell carci- 
nomas of the skin being predominatingly in old mice, and as 
many male mice die early from wounds received in fighting 



70 MAUD SLYE, H. F. HOLMES AND H. G. WELLS 

this may account for the relative preponderance of females in 
this series. 

Beyond the relatively slight extent of infiltration and metas- 
tasis these carcinomas present no noteworthy differences from 
human skin carcinomas. Often the amount of keratin scales piled 
up on the surface is strikingly great, and in one case the carci- 
noma arose at the base of definite cutaneous horns (25785)* 
There is often a noteworthy aniount of calcification of the 
necrotic scales. Cyst formation is frequently observed, and 
benign cutaneous cysts have been observed in several mice. 

The basal cell carcinomas all arose on the face, ears, and head 
(fig. 1), 9 of the 15 being in females. Because these growths 
are of relatively low malignancy, as in man, they present partic- 
ular diflSculty in diagnosis, and we have set aside as '^precan- 
cerous'' numerous instances of basal cell hyperplasia of marked 
degree which did not present unqualified proof of malignancy. 
Quite frequently enormous hyperplasia of the cells about the 
hair follicles produces tumors of considerable size, and in these 
may be found areas highly suggestive of maUgnancy, in that 
altogether atypical plugs and masses of basal cells are formed, 
as shown in figure 6. Such growths, which very probably 
would have shown unqualified malignancy had the mouse lived 
longer, have not been included among the basal cell carcinomas. 
In general, basal cell carcinoma in the mouse corresponds entirely 
to the corresponding growth in man (fig. 7). To quote Mac- 
Callum (28), ''In spite of the complexity of the downward 
growing strands all reach to about the same level. Further, 
it is seen that they are very sharply outUned against the stroma 
and show little inclination to strew their cells into the irregular 
crevices of that tissue. '' We have found the various modifica- 
tions of basal cell growths commonly described, such as the 
formation of tubules suggestive of undeveloped hair follicles, 
homifyijig surface areas with typical basal cell growths beneath, 
and the so-called adenoid epithelioma which seems intermediate 
between basal and squamous cell growths. Among our keratin- 
izing carcinomas of the mammary gland was one of cells 
that suggested a basal cell character. No metastases were 



FiQ. 6. Basal Cell HvPERpLAaiA or an* Ihritated Muzzle, with Areas of 
More Atypical Growth 
There have been numerous auch growths, which are probably properly desig- 
nated as precancerous, but may be actually malignant. No. 16663. X 60. 



o. 7. Basal Cell Carcinoma Which was Highly Malignant, as '. 

Entirely Replaced tub Face, Destroying Both Eyes, andForuino 

Tumor Masses in- the Orbits. No. 9932. X 90 



72 MAUD SLYE, H. F. HOLMES AND H. G. WELLS 

observed among these cases, although in one, extensive infiltra- 
tion destroyed both eyes and produced large tumor masses in 
the eye sockets. Several of these growths arose at the site of 
distinct scars from wounds. 

Despite the fact that these skin tumor mice were all well 
advanced in age the co-existence of other tumors is lower than 
that seen in many other forms of mouse tumor. Perhaps one 
factor in this is that most of the mice with skin cancer had lived 
beyond the age at which other tumors occur most frequently. 
It also seems that the heredity of these mice is responsible to 
some extent, but a complete analysis of this factor has not yet 
been made. Of the 85 cases, in which 58 were females and 27 
males, but 6 showed a carcinoma of the mammary gland, 7 
exhibited lung adenoma (including one that also had a mammary 
gland carcinoma), one had an adenoma of the ovary and one 
seemed to have pseudoleukemia. Despite the number of old 
wounds in these mice no cases of sarcoma were observed, presum- 
ably because of either old age or ancestry, but one case of carci- 
noma arising in the mouth and infiltrating the jawbone, showed 
such a flattening of the deeper cells that for some time we were 
in doubt whether or not it was an instance of mixed sarcoma 
and carcinoma. This case (8560) is described and pictured in 
our article on sarcoma in mice (29), 

SQUAMOUS CELL CARCINOMAS OF THE MAMMARY GLAND 

The keratinizing carcinomas of the mammary gland form an 
interesting group, and properly lie outside the scope of this 
paper, except for the fact that with not a few of them it is 
extremely diflBcult to tell whether we are dealing with a primary 
mammary gland carcinoma or a skin carcinoma arising over the 
mammary gland. Three types of these tumors can be dis- 
tinguished. 

1, Carcinomas of the mammary gland which are essentially 
cylindrical cell carcinomas, forming tubules and alveoli, but 
some areas of which imdergo a transformation into stratified 
epithelium with the formation of keratin, often in large amounts 
(fig. 8). These are the " adenocancroids " of Murray and 



Fig, 8. Tdbular Alveolar Carcinoma ("Adbnocancroid") of the Mammary 
Gland with Areas of Keratinization 
To the left is Been a band of densely keratJDized tissue with no evidence of 
its glandular origin. No. 13336. X fiO. 



FiQ. 9. Superficial Adenocancroid, Apparently Arisino in the Ducts of 
THE Mahuarv Gland 
Shows very little evidence of origin in glandular tissue. This mouse also 
had three other tubular carcinomas of the mammary gland without keratiniza- 
tion No. 15300 X 60. 

73 



74 MAUD 8LYE, H. F. HOLMES AND H. G. WELLS 

Haaland, It is probably of some significance that the keratin- 
ization is usually most marked in the portions nearest the 
cutaneous surface, as if it began in the ducts. The amount of 
keratinization varies, sometimes appearing in only a few small 
spots in the tumor tissue, but often extending until little of the 
original columnar cell type of tissue remains. 



Fia. 10. Keratinisino Cvstb in tub Mamuart Gland 
Such structures are not infrequently found, and may be the precursors of some 
of the malignant keratinizing carcinomas of the mammary gland. No. 8271. 
X40. 

2, Squamous-celled keratinizing carcinomas without any 
evidence whatever of cylindrical cell structure, but arising 
subcutaneously in the manimary regions (fig. 9). Many of these 
present no histological evidence that they are derived from the 
cells of the mammary gland, but they are observed to arise 
within the gland substance and sometimes are still entirely 
subcutaneous when the mouse dies. The occasional presence 



SQUAMOUS CELL CARCINOMAS IN MICE 75 

within the mammary gland of what seem to be simple benign 
cysts of stratified epithelium with masses of desquamated homi- 
fied material (fig, 10) indicates the probable origin of such tumors. 
When they ulcerate on the surface it may be impossible to 
differentiate them from primary carcinoma of the skin or nipple, 
for histologically they differ little if at all from the usual squamous 



Fia. II. MoLLuscoiD Carcinoma of the Mauuart Gland 
This growth, which arose in an inguinal mammary gland, invaded the abdomi- 
nal wall and protruded into the pelvis, but produced no metastases. A non- 
keratinizing carcinoma of the mammary gland was also present which produced 
pulmonary metaetaees. No. 15622. X 60. 

cell carcinoma of the skin. When observed early their origin 
in the gland beneath the skin is usually the chief ground for 
recognizing them as mammary gland tumors; nevertheless, the 
fact that after differentiating, on a histological basis solely, the 
skin carcinomas from the mammary gland carcinomas we found 
that nearly all those we had selected as of cutaneous origin had 



76 MAUD SLYE, H. F. HOLMES AND H. G. WELLS 

arisen at the site of old skin wounds, indicates that there are 
usually recognizable differences. 

3. MoUuscoid carcinomas of the mammary gland, to adopt 
Borrel's designation of these striking growths with long, radia- 
ting cylinders of stratified epitheliun^ with a large central core of 
keratinized material (fig. 11). These seem to grow out from the 
tubules near the nipple, as if they arose from the stratified 
portion of those ducts. They produce ulceration of the surface 
about the nipple, which greatly enhances their resemblance to 
the skin carcinomas. 

We have little to add to the statements of Borrel, Murray, 
Tyzzer, and Haaland concerning these squamous cell growths 
arising within or about the mammary gland, and the excellent 
illustrations accompanying Haaland's article make any further 
illustrations by us unnecessary. A few points may be worth 
recording. Forty-three of the 56 squamous-celled manamary 
gland tumors may be best designated as of the adenocancroid 
type, in that they show a greater or less proportion of cylindrical 
cell structure. Several of these were multiple, as is often the 
case with primary carcinoma of the mammary gland. In three 
there were two adenocancroids, one mouse had three, and one 
had five, four arising distinctly in mammary gland tissue while 
the fifth arose anterior to the urethra and seemed to have origi- 
nated in the skin, and this showed no glandular elements. Four- 
teen mice had an* adenocancroid and also one or more other 
independent mammary gland carcinomas that showed no keratin- 
ization. Of these, one (9365) exhibited also a typical squamous 
cell carcinoma metastasis in the lung; one mouse with three 
primary adenocancroids (5738) showed a lung metastasis while 
three mice that had each two mammary gland tumors, one of 
each type, had pulmonary metastasis of the cylindrical cell type. 
That is, from 43 cases of adenocancroid of the mammary gland, 
but one metastatic nodule of adenocancroid has been seen in 
the lungs, and none in the lymph nodes or elsewhere. Not a 
few of these mice had also co-existing tumors other than those of 
the mammary gland, namely, one with a squamous cell carci- 
noma of the muzzle (9962), one with a malignant ovarian tumor 



SQUAMOUS CELL CARCINOMAS IN MICE 77 

(6801), two with uterine fibroids (14370 and 14811); one with a 
lung adenoma (12098) ; one with an adenoma of the hver and an 
adenoma of the limg (9544) ; one with an adenoma in the limg 
and a strange growth in a subcutaneous lymph node that might 
be interpreted as an endothelioma (15622) and one had an intra- 
pelvic sarcoma with metastasis into the liver and also into an 
ovary which contained in addition a small adenoma (10006). 

Two of the adenocancroids were examples of the not uncommon 
mixture of sarcoma and carcinoma in the mammary gland of the 
mouse (15412 and 6182). 

Only four of our tumors were typical of the molluscoid cancers, 
although more or less of this character was shown by several of 
the growths diagnosed as adenocancroids. One of these four 
mice had also an adenoma of the ovary (496) and another had a 
benign limg adenoma (5581). None of these four keratinizing 
tumors produced metastases. 

Nine squamous cell carcinomas that exhibited no glandular 
structure to identify them as mammary gland tumors, were 
believed to have this origin because they arose beneath the skin, 
and in most instances did not ulcerate through at any time. 
These growths, however, seem to be of a more malignant character 
than the adenocancroids; three of them produced squamous 
cell metastases. No. 10936, which is remarkable in being the 
only male with a subcutaneous stratified epithelial growth that 
seemed to arise in the mammary gland, exhibited squamous 
cell metastatic growths in the lung, in the mediastinum, and in the 
chest waU. Another (7950), with three subcutaneous, non- 
glandular squamous cell carcinomas, showed a metastatic nodule 
in one Ijonph node and direct infiltration of a second. No. 
13671 with a subcutaneous squamous cell carcinoma and also a 
cylindrical cell mammary gland carcinoma, had three large 
metastatic nodules in the lungs all of squamous cell structure. 
As to multiplicity of tumors in this group there were three 
independent subcutaneous squamous cell tumors in each of two 
mice (2239 and 7905), two had a simple cylindrical cell carcinoma 
of the mammary gland in addition to the squamous cell growth 
(7526 and 13671), while one had adenomas of the liver and ovary 



78 MAUD SLYE, H. F. HOLMES AND H. G. WELLS 

(14047) and one mouse had leukemia (8236). In this group 
we have dncluded one squamous cell tumor (24858) that differs 
from the rest in having the non-keratinized tumor elements 
suggest strongly the appearance of a basal skin carcinoma^ 
although the growth seemed to arise in the manunary gland. 

SQUAMOUS CELL CARCINOMA OF STOMACH AND RECTUM 

In a previous paper on primary carcmoma of the stomach 
in mice (27) we collected records of four reported cases of squa- 
mous cell carcinoma arising in the cardiac portion of the stomach 
to which we added three observed in this laboratory. Since 
that publication we have observed one more case (24367). This 
was an old female mouse that for some time before death had 
been isolated because suffering from tape worm infection. The 
stomach wall in the cardiac portion was much thickened and 
small white nodules were scattered over the outer surface. No 
metastases could be found. In the liver was an encysted cestode, 
and there was a prolapsed rectum showing much thickening of 
the wall. Microscopically the nodules in the stomach are of 
squamous cell carcmoma^ apparently still quite early but of 
typical structure. One small nodule of sunilar structure is foimd 
in the omentum. The rectum shows a marked overgrowth of 
squamous epithelial plugs on the outer, ulcerated surface of the 
bowel, and with many of these an infiltrative character is sug- 
gested, but apparently this lesion cannot be diagnosed as 
carcinoma. 

Another new case of carcinoma of the stomach (25911) is more 
difficult to classify. A female Peromyscus mouse, nearly four 
and a half years old, which had been isolated all her life and 
never bred, refused food for several days before death. At 
autopsy there was found a marked thickenmg of the entire 
stomach wall, particularly at the cardiac end where it seemed 
nearly solid from wall to wall. A thick white nodular outgrowth 
6 by 4 by 4 mm. of similar tissue bound together the stomachy 
liver, and esophagus, while another nodule 8 by 4 by 4 mm. 
boimd together the pancreas and intestinal mesentery posterior 



SQUAMOUS CELL CARCINOMAS IN MICE 79 

to the stomach. No other metastases were fomid. Microscop- 
ically this growth is composed of groups of strands of infiltrating 
epithelial cells, which are not homified, neither do they form 
tubules. They grow profusely throughout all coats of the 
stomach, sometimes forming sheaths about the tubules of the 
cardiac portion. The nodules outside the stomach are secondary 
growths of identically the same structure in lymph nodes. It 
is not possible to be sure whether this carcinoma is derived from 
the columnar or the stratified epithelial portions of the stomach. 
While the gross appearances suggested an origin in the cardia, 
which is the site of all but one of the recorded cases of gastric 
carcinoma in mice, yet all of these were frankly homifying 
squamous cell carcinomas. 

The full details of the three other cases of squamous cell 
carcinoma occurring in this series are published with illustrations 
in our previous communication (27) and need not be repeated 
here. To recapitulate them briefly they were as follows : 

No. 5802, male two years old. Squamous cell carcinoma of 
pyloric portion of the stomach, measuring 15 by 15 by 12 mm. 
A secondary growth 11 by 12 mm. in the mesentery. No other 
metastases. Some infiltration of the pancreas. 

No. 7851. Male, twenty-five months old. Cardiac portion 
of the stomach is ulcerated and thickened, forming a mass 16 
by 12 by 8 mm. No metastasis. Structure, typical squamous 
carcinoma, infiltrating all coats of the stomach, and invading 
the adhesions between the stomach and liver. 

No. 16440. Female, aged twenty-seven months. At the 
jimction of the cardiac and pyloric portions a thickened mass 
from 3 to 15 mm. wide surrounds the stomach. Nodules were 
found in the omentum and mesentery, metastases in lymph 
nodes. Structure, typical squamous cell carcinoma with some 
homification. 

In the same paper was published with illustrations the report 
of a case of squamous cell carcinoma arising in the metaplastic 
epithelium covering a prolapsed rectmn (8345). This was a male 
mouse, which had had a prolapsed rectum for six months before 
its death. The growth was not extensive but seemed to be 



80 MAUD SLYE, H. F. HOLMES AND H. G. WELLS 

typically squamous cell carcinoma arising in metaplastic epi- 
thelium. Since then we have observed a second similar case. 

No. 20052. Female, had a prolapsed rectum for four months 
before death, which was from h3T)ertrophy of the heart and 
pulmonary edema. The surface of the prolapse became much 
ulcerated and after a time showed what seemed to be a prolif- 
eration of epithelimn. The regional l3ntnph nodes were some- 
what enlarged. Microscopically the external surface of the 
prolapsed bowel shows, in addition to considerable ulceration, 
areas of definite squamous cell carcinoma, with strands of cancer 
cells infiltrating through a thick layer of granulation tissue 
down to but not into the muscularis. Apparently this has 
arisen from what was part of the cutaneous surface of the anus, 
and not from metaplastic mucosa as in the previous case. As 
a possible factor in the inciting irritation, a small piece of wood 
was foimd imbedded in the bowel waU, surrounded by granula- 
tion tissue and cancer cells. 

CARCINOMA OP THE VULVA AND VAGINA 

No cases of squamous cell carcinoma of the uterus have as 
yet been described in mice, so far as we can learn. Although 
we have found a few fibromyomas and sarcomas of the uterus, 
we have met with but one epithelial neoplasm, which was an 
adenocarcinoma. 

Erdheim (23) and Haaland (20) each have described a recurrent 
verrucous growth of the vulva, and the latter a mixed squamous 
and sebaceous cell carcinoma of the vulva. Two cases of 
squamous carcinoma of the vulva have been observed in this 
series and are included among the 70 squamous cell skin carci- 
nomas. (1) No. 7950. This occurred as a condylomatous 
growth arising distinctly in the vulva of an old mouse, infiltra- 
ting the subcutaneous tissues, and was an unquestionably 
malignant squamous cell growth microscopically. There was in 
addition metastatic growth in two lymph nodes in the groin. 
(2) No. 18928. A growth of a warty character and ulcerating 
slightly developed on the vulva some time before the death 



SQUAMOUS CELL CARCINOMAS IN MICE 81 

of the mouse, spreading about the rectum before death. At 
autopsy there was found m addition an enlarged Isrmph node 
attached to the right ureter. Microscopically the external 
growth is a typical squamous cell carcinoma with much homi- 
fication, and not very marked tendency to infiltration. The 
nodule attached to the ureter was a Ijrmph nonde containing a 
cyst lined with squamous epithelium, or, to describe it better,, 
an epithelial cyst covered with a thin layer of lymphoid tissue. 
The epithelial wall of the cyst is thin, shows no evidence of infil- 
trative or other mahgnant character, and hence does not at all 
resemble a metastatic growth, but it is difficult to explain the ' 
presence of such a structure in this location on any other basis. 

In addition to these two carcinomas we have observed an 
excellent case of carcinoma of the vagina. 

Carcinoma of vagina (no. 22582). This mouse was foimd, 
nearly a month before its death, with a 3 mm. pink nodule 
protruding from the vagina. At the time of its death from 
pneiunonia it presented an ulcerating mass 15 by 12 by 10 mm., 
about one-half of which protruded from the vulva. It seemed 
to arise from the vaginal wall, the uterus and bladder not being 
involved, but it ulcerated into the rectum and was much infected 
and ulcerated. No enlarged glands or other evidences of metas- 
tasis could be found. Microscopically the growth is composed 
mostly of loose masses of keratinized scales exfoliated from the 
underljdng growth, which infiltrates the vaginal wall as a typical 
squamous cell carcinoma. 

KERATINIZING TUMORS OF THE LUNG 

As mentioned in our review of the literature, Haaland has 
described one case of primary keratinizing growth in the lung, 
and Tyzzer says that he has had "several cases. " The influence 
of heredity on the incidence of tumors of special types is suggested 
by the fact that of several himdred cases of primary lung tiunors 
in the Slye stock, but. one has been definitely found to show 
keratinization. 

No. 13314. A male, age twenty months, which showed no 
other autopsy findings of interest, had the lower lobe of the left 



82 MAUD SLYE, H. F. HOLMES AND H. G. WELLS 

lung nearly replaced by a yellowish mass, which was distinguished 
from the ordinary papillary adenoma of the lung chiefly by its 
color. This made it resemble an abscess, but it was much too 
hard to be an abscess, and there was no pleural exudate or 
adhesion. Microscopically the tumor consists chiefly of a mass 
of homified scales heaped up in waving, concentric layers. 
Only at the very periphery are living cells found. Here is a 
narrow growing border of stratified epithelial cells, differing in 
no essentials from that seen in epitheliomas of the skin. There 
is no marked tendency to infiltration, the growth apparently 
progressmg by expansion, but the presence of occasional mitotic 
figures is noted. About the growth there is much round and 
spindle cell proliferation and numerous foreign body giant cells 
jape dfound. Our specimen differs from the one illustrated by 
Tyzzer in having a smaller proportion of living cells. We are 
not certain whether this growth represents a true neoplasm or a 
progressive metaplasia due to some persisting chronic inflam- 
matory condition, but the former seems more probable. 

Two other mice have shown somewhat related pulmonary 
conditions. 

No. 10561. This mouse, which had also a carcinoma of the 
skin, had a benign adenomatous growth in the lung which showed 
some tendency to stratification, but without keratinization. 

No. 25136. A male mouse had in the right upper lobe of the 
lung a mass 14 by 12 by 10 mm. resembling in appearance a 
malignant tmnor of the lung. Microscopically this tissue 
resembles much more closely an unresolved organizmg and 
necrotizing pneumonia, in which are two irregular, independent 
areas composed of masses of keratmized scales, with a slender 
border of flattened epithelial cells. It is quite impossible to 
decide whether this is a true tumor, or whether it is the cause or 
the result of the pneiunonic condition, although the epithelial 
growth has the appearance of being much older than the pneu- 
monic process. 



SQUAMOUS CELL CARCINOMAS IN MICE 83 

TUMORS OF THE MEIBOMIAN GLAND 

Two mice have presented growths arising in the eyelids, 
which, according to their structure, seem to be adenomas arising 
in the Meibomian glands. 

No. 18099. A small, slowly growing mass developed beneath 
the left eye of a female mouse; from the eye exuded a small 
amount of thick white exudate. The growth had reached a 
diameter of 10 mm. when the mouse died from an acute lung 
infection, and showed no evidence of infiltration or ulceration. 
Microscopically the growth is composed of papillary structures 
covered with many layers of epithelial cells. It differs from the 
normal Meibomian gland in the exaggeration and lawless arrange- 
ment of the structures, and the greatly increased number of 
epithelial cells covering the stroma. The diagnosis of benign 
adenoma of the Meibomian gland seems justified, especially in 
view of the size of the growth and the findings in the next case. 

No. 27929. An old female mouse, which died of chronic 
nephritis, had a soft mass, 6 mm. in diameter, beneath the left 
eye. The gross appearance suggested an epithelioma. Micro- 
scopically this tumor is quite the same as the one described 
immediately above, except for the important fact that it infil- 
trates down to the bone of the orbit, thus indicating that it is a 
malignant infiltrating adenocarcinoma. 

SEBACEOUS GLAND ADENOCARCINOMA 

Such tumors have been described in mice by Murray, Tyzzer, 
and Haaland, the last two having transplanted them success- 
fully. We have found one case of sebaceous adenocarcinoma of 
the preputial gland, which closely resembles the growth described 
and illustrated by Haaland (20). 

No. 19895. Beginning at the base of the penis is a mass 30 
by 25 by 25 mm. extending well into the inguinal region. The 
penis was completely imbedded in the tumor, the older portions 
of which were softened, but about the periphery were hard 
nodules of newer growth. The testicles and epidid3anis were 



84 MAUD SLYE, H. F, HOLMES AND H. G. WELLS 

not involved. Death resulted from chronic nephritis. There 
were no metastases. Microscopically this tumor (fig. 12) repro- 
duces closely the normal structure of the preputial gland, but 
not infrequent infiltration of the stroma by strands of epithelial 
cells corroborates the gross evidences of malignancy. It corre- 
sponds perfectly to the illustration given by Haaland, who also 



Fio. 12. Sebacbods Ade no-carcinoma of Preputial Gland 
This was a Urge tumor with distmctly malignant characteriatics, infiltrating 
the adjacent tissues, despite its close resemblance to the normal struoture of 
the gland from which it arose. No. 10895. X 60. 

comments on the close resemblance to the normal gland in spite 
of the definitely malignant character. 



Among 28,000 mice dying natural deaths at all ages, and 
carefully autopsied, have been observed 153 growths of strati- 
fied and squamous epithelium that correspond by the usual 



SQUAMOUS CELL CARCINOMAS IN MICE 85 

standards to true neoplasms, excluding a considerable number 
of epithelial growths which lack positive conclusive evidence of 
neoplastic character, although possibly some of these are also 
early carcinomas. Seventy-one are examples of squamous cell 
carcinoma of the skin or mouth. They differ from the human 
skin carcinoma chiefly in a low incidence of metastasis. Fifteen 
others are of basal cell character, arose always about the head, 
and produced no metastases. In both these groups the inicidence 
is higher in the females than in the males. Trauma and chronic 
irritation seem to play an important part in the production of 
skin carcinoma in mice, most of our cases occurring about the 
head and face, often recognizably at the site of wounds, and 
nearly all the skin carcinomas of the trunk arose in old scars. 
Skin cancers occur at a greater average age than other tumors 
in mice. 

Fifty-six examples of squamous cell keratinizing growths 
arising in the mammary gland were observed, predominatingly 
adenocarcinomas with localized areas of keratinization. These 
also seldom produce squamous cell metastases. 

Other tumors in this group were: Four squamous cell carci- 
nomas of the stomach, two arising in the prolapsed rectum, two 
in the vulva, one keratinizing tumor of the lung, one sebaceous 
adenocarcinoma of the preputial gland, and, as hitherto imde- 
scribed mouse tumors, one squamous carcinoma of the vagina 
and two adenomas of the Meibomian glands, one of these being 
infiltrative and apparently malignant. 

The literature of the comparative pathology of squamous 
cell carcinoma in animals is reviewed, and it is worthy of comment 
that as yet no cases of squamous cell carcinoma of the uterus, 
bladder, or esophagus seem to have been described in mice. 

REFERENCES 

(1) Sticker: Arch. f. klin. Chir., 1902, Ixv, 616, 1023. 

(2) Teutschlaender: Ztschr. f. Krebsforsch., 1920, xvii, 285. 

(3) JoEST AND Ernesti: Ztschr. f. Krebsforsch., 1915, xv, 1. 

(4) Fox: Jour. Path, and BacterioL, 1912, xvii, 217. 

{5) Botnton: Festschrift for Victor C. Vaughan, 1903, p. 635. 
(6) Urra: Rev. de med. y cirug. prdct., Madrid, 1918, cxviii, 321. 



86 



MAUD 8LYE, H. F. HOIjMES AND H. G. WELLS 



C 



(7: 

(8] 

(9] 
(10) 



'-^ 



^^ (u: 

^12] 
(18] 
(14] 
(15] 
(16] 
(17] 
(18] 
(19] 

(20] 

(2i; 

(22] 
(23] 
(24] 
(25] 
(26] 
(27] 
(28] 
(29] 
(30] 

(3i; 



Tbutbghlabnder: Ztschr. f. Krebsforsch., 1920, xviii 350: 

Fiebioeb: Ztschr. f. Krebsforsch., 1909, vil, 105. 

RoFFo: Rev. d. Inst. Bacteriol., Buenos Aires, 1918, 1, 333. 

Bashford: Third Sci. Rep.j Imperial Cancer Research Fund, London, 

1908, p. 41. 
Loeb: Arch. f. klin. Chir., 1903, ]xx, 845. 
WooLtiET AND Whxbbt: Jour. Med. Res., 1911, xxv, 205. 
Hanau: Fortschr. d. Med., 1889, vii, 321. 
RoFFo: Rev. d. Inst. Bacteriol., Buenos Aires, 1918, i, 155. 
LipschVtz: Wien. klin. Wchnschr., 1920, xxxiii, 426. 
Bullock and Rohdenbtjro: Jour. Cancer Res., 1917, ii, 39. 
Morris: Jour. Cancer Res., 1920, v, 147. 
Haaland: Ann. de Tlnst. Pasteur, 1905, xix, 165. 
Murray: Third Sci. Rep., Imperial Cancer Research Fund, London, 1906, 

p. 69. 
Haaland: Fourth Sci. Rep., Imperial Cancer Research Fund, London, 

1911, p. 1. 
Tsutschlabndbr: Ztschr. f. Krebsforsch., 1920, zvii, 300. 
Pick and Poll: Berl. klin. Wchsnchr., 1903, xl, 518. 
Erdhbim: Ztschr. f. Krebsforsch., 1906, iv, 33. 
TsuTBUi: Gann, 1918, xii, 17. 
Ttszbr: Jour. Med. Res., 1909, zxi, 479. 
JoBLiNo: Rockefeller Inst. Monographs, 1910, i, 81. 
Sltb, Holmes and Wells: Jour. Cancer Res., 1917, ii, 401. 
MacCallum : Text-book of Pathology, Philadelphia, 1920, p. 1032. 
Slte, Holmes, and Wells: Jour. Cancer Res., 1917, ii, 1. Fig. 3. 
Woglom: Jour. Cancer Res., 1917, ii, 471. 
Fiebioeb: Jour. Cancer Res., 1919, iv, 367. 



PROCEEDINGS OF THE AMERICAN ASSOCIATION FOR 

CANCER RESEARCH 

THIRTEENTH ANNUAL MEETING 
Held in New York, April 1, 1920 

1. Report of the Council 

The meeting of the Council was held at Dr. Wood's apartment in 
New York City, on the evening of Wednesday, March 31, 1920. 

The following members were present: Dr. H. Gideon Wells, presi- 
dent; Dr. Robert B. Greenough, vice-president; Dr. Francis C. Wood, 
and Dr. Willy Meyer, and, by invitation Dr. Frederick Prime. Absent, 
Dr. E. R. LeCoimt, Dr. James B. Murphy, and Dr. William H. Woglom. 

The report of the treasurer for the year showed a balance on hand 
of $532.27. 

The status of the Joubnal of Cancer Reseabch was discussed, 
and a statement from the pubUshers, showing a total deficit of $110.80 
on the four volumes so far published, was re£ul. After some discussion 
of the proposal, made by the publishers, that the price of the Joubnal 
be increased to $6.00 for subscribers not members of the Association, 
it was voted that no change be made until after the completion of the 
pubUcation of Volimie V. 

In order to increase the membership of the Society it was decided 
that the president should send out a letter to certain men whose 
names were presented as possible candidates for active and associate 
membership. 

The name of the following applicant came before the Coimcil for 
election to the Association: 

Dr. Michael Levine, Montefiore Home, New York City. 

Dr. Wood moved and Dr. Wells seconded the motion that he be 
elected to membership in the Association. 

The resignations of the following members were accepted: 

Dr. J. George Adami, Dr. Martha Tracy. 

The deaths of Dr. M. J. Herzog and Dr. James Douglas were' noted. 

Dr. Wood moved that Dr. James Ewing be elected councillor suc- 
ceeding the retiring coimcillor. Dr. Greenough seconded this motion. 

The following officers were elected by the Coimcil to serve for the 
ensuing year. Dr. Robert B. Greenough, president; Dr. James B. 
Murphy, vice-president; Dr. WiUiam H. Woglom, secretary and 
treasurer (re-elected). 

87 

THB JOUBKAL OT CAMCXR BUEABCB, TOL. .▼!, KO. 1 



88 PROCEEDINQS 

The present Council, therefore, with the years of retirement, is as 
follows: 

Dr. E. R. LeCount, 1921 Dr. William H. Woglom, 1924 

Dr. F. C. Wood, 1922 Dr. Robert B. Greenough, 1925 

Dr. James B. Murphy, 1923 Dr. Willy Meyer, 1926 

Dr. James Ewing, 1927 

The Council continued in office the present Editorial Board, which 
is composed as follows: 

Dr. William H. Woglom Dr. Leo Loeb 

Columbia University Washington University 

Dr. Frederick Prime Dr. Eniest E. Tyzser 

Columbia University Harvard University 

Dr. Joseph C. Blood{;ood Dr. H. Gideon Wells 

Johns Hopkins University University of Chicago 

Dr. James Ewing 
Cornell University 

2. A Few Late Results Afteb the Radical Operation for 

Cancer of the Breast 

Dr. Willy Meyer (New York) : 

SUMMARY 

The data given are from the author's private cases which he has 
followed for the last twenty-six years. 

Two radical operations for cancer of the breast have been before the 
profession since the fall of 1894. Their principal point of difference is 
the direction in which the surgeon proceeds. One method starts from 
the chest and works toward the axilla, leaving the clavicular portion 
of the pectoralis major behind; it requires entering the space between 
the pectoralis major and minor muscles, the latter usually^ being 
divided and then sutured. This method necessarily involves loss of 
blood. The other method, practised by the author since September 
12, 1892, starts from the axilla and works toward the sternum. The 
tendons of the pectoralis major and minor are divided in the early 
stage of the operation, necessitating complete excision of both muscles. 
Bl(K>d- and lymph-vessels are primarily divided within the axilla. 
The lymph-nodes and axillary fat are Ufted out in connection with 
the tumor, before the cancerous breast itself is handled. The entire 
mass is removed without entering what he calls the "infected area." 
Hemorrhage is reduced to a minimum. 

The final results of the operation from the sternum toward the 
shoulder, as reported, have been good. Still, small cancerous nodes 
have been found repeatedly between the pectoralis major and minor 
muscle, and where ccmcerous l3rmphatic nodes have developed, there 
must be present suspicious lymphatic vessels. 



PROCEEDINGS 89 

Previous to 1894, excision of the breast for carcinoma was done in 
two stages, but at the same sitting, first, the removal of the breast with 
axillary contents; then, the excision of the pectoraUs major muscle. 
This arrangement forced the surgeon to enter widely ''the infected 
area" and caused an mmecessarily great loss of blood. Meyer did not 
see a single lasting cure after this method of advance. The radical 
operation changed the results with one stroke, as is shown by a com- 
parison of the author's statistics before and after September, 1894. 
The first two patients subjected to the modem radical operation were 
completely cured, and case 4 of this series had for many years enjoyed 
freedom from cancerous recurrence when she died of old age. 

The following patients in the series are alive and well today, from 
twelve to twenty-five and a half years after operation: 

Case 1. Operation in September, 1894; patient thirty-eight yean of age 
(now sixty-four). This was the first case operated upon by the method out- 
lined above. The patient is alive and well today, twenty-five and a half years 
after operation. 

Ciue i. Operation in 1895; patient forty-eight years of age (now seventy- 
three). This woman is alive and well today, twentv-five years after operation. 

Case 5. Operation in July, 1902: patient thirty-three years of age (now fifty- 
one). The patient is alive and well today, eighteen years after operation. 

Case 4' Operation in December, 1903; patient thirty-six years of age (now 
fifty-three). The patient is perfectly healthy and free from recurrence today, 
seventeen years after operation. 

All these patients have full use of the arm and are able to assume the posture 
of the "Statue of Liberty." 

Case 5. Operation in July, 1908; patient thirty-five vears of age (now forty- 
seven). She IS perfectly healthy toaav, twelve years after operation. 

Case 6, This patient was operated on in September, 1917, and the case is 
added merely^ to show the present line of incision with Handley's addition down 
to a point midway between the umbilicus and the xyphoid process, for the ex- 
cision of the fascia covering the upper portion of the recti muscles, in conjunc- 
tion with the other mass. This addition is recommended because it makes the 
operation more radical and usually makes possible closure of the wound, without 
grafting. 

Five other patients remained free from recurrence for 4, 6, 8 (2 
cases), and 16 years, respectively, and then died of other diseases. 

Another patient, a pronomiced diabetic at the time of the operation, 
was well for six years after it, when she succmnbed to diabetes, without 
having developed any signs of a recurrence of the cancer. 

Still another patient, operated on for cancer of the right breast in 
March, 1899, returned in December, 1900, with a carcinoma of the 
left breast, which was then also extirpated. She was well and free 
from recurrence when last heard from, in the spring of 1907, six and a 
quarter years after the second operation. 

A few ddkjs ago, the author saw a patient now almost eighty years 
old and in perfect health, who had been operated upon by him for a 
scirrhus of the breast at the age of seventy-three (seven years ago). 

These results, he beheves, prove the efficiency of the method; they 
prove that the radical operation for cancer of the breast can cure patients 



90 PROCBBDINGft 

thus afflicted. That not all tlhe patients are saved is due (1) to the 
stage of the disease in which they reach the surgeon, and (2) to the 
virulence of the agent that produces the carcinoma. 

The importance of the follow-up system was discussed in connection 
with the collection of statistics on the late results of operation. 

Pagefs disease (epithelioma of the nipple) 

Paget's disease is the most malignant of known forms of cancers of 
the breast. If ever early and radical operation is imperative, it is in 
these cases, as will be seen from the f ollowii^ three observations which 
were made diuing the last two years. 

Case 1, Female, thirty years old, mother of five children, had been in the 
hands of a quack and had been treated by caustics. When seen in January, 1918, 
the disease in the breast and axilla had far advanced. After the radical oi>era- 
tion the other breast soon became affected, and one year later, also was excised. 
Then, not long after, the disease became disseminated, and the patient died 
from cancer en cuirasse. 

Case M, Female, thirty-eig^ht yean old, had been in the hands of one of our 
best x-ray specialists- in the city. One and one-half vears after cure of the dis- 
ease of the nipple by radium treatment, there was a local recurrence and a very 
extensive cancer of the breast, with infected nodes in the axilla and along the 
subclavian vein. Radical operation was done in December, 1918, followed by 
renewed x-ray and radium treatment. She now has developed intrathoracic 
metastases. 

Case 5. Male, forty-seven years old, had been in the hands of an experienced 
surgeon who had extirpated the breast only, without the axillary nodes, evidently 
because none could be found at that time. One and three-quarter years later, 
the patient presented a far advanced carcinoma. The radical operation then 
performed could not save him; and he died from general metastases eight months 
afterward. 

In operating upon mammary carcinoma, Meyer makes it a point to 
circmncise the skin widely at the base of the breast. He prepares two 
ample flaps and enfolds them extensively, then divides the fasciae at 
the base of the two flaps, and extirpates them together with the mass. 
Involvement of the supraclavicular nodes does not, in his opinion, 
present a contraindication to operation; on the contrary, he considers 
it the surgeon's duty to operate when these nodes are infected. Hal- 
st^, as well as the late Rodman, have observed patients who remained 
well for a number of years after the extirpation of these nodes. 

In none of the cases presented were the supraclavicular nodes found 
infected at the time of operation, and hence they were not removed. 
Meyer has not operated on a single case in which there were no infil- 
trated axillaiy nodes. 

Meyer believes that statistics regarding the results of the radical 
operation for cancer of the breast are worthless. They do not prove 
anything. What does determine the fate of the patients is the so- 
caUed virulence of the disease. One and the same surgeon may do an 
equally radical operation in two seemingly early or apparently equally 
far advanced cases, in one of which the patient may remain well and 



PROCEEDINGS 91 

free from recurrence for, say, twenty-five years, while in the other a 
regional recurrence and metastases may develop within a few months. 
All that can be said is that cancer, being a local (lisease in the begin- 
ning, may be cured by a radical operation, if this is done at an early 
stage. 

DISCUSSION 

Dr. Robert B, Greenough (Boston) ; I feel we must acknowledge that 
surgery of the breast owes much to Dr. Meyer for the work he has 
been doing, and I am entirely in accord with his reasons for operating 
in the way which he recommends, namely,* beginning the operation in 
the axilla, and removing the axillary contents, the whole of the breast, 
both of the pectoral muscles, and the fascia of the chest wall, all in one 
piece. I feel this hardly needs argument, because it is an accepted 
principle in operating for the radical cure of cancer in any situation 
where the anatomical conditions make it possible. I personally have 
had more experience with the transverse axillary incision than with 
the incision used by Dr. Meyer, but I think the choice of the incision is 
a very small matter. It is far more important what is done, beneath 
the skin, and I am entirely in accord with Dr. Meyer in that respect. 

In regard to the presentation of statistics in cases operated on for 
the cure of cancer, some fourteen years ago I had the opportunity of 
reporting, at the meeting of the American Surgical Association, the 
results of operations for carcinoma of the breast at the Massachusetts 
General Hospital, and at the same meeting other papers on the end 
results of cancer at many different institutions were also presented. 
Dr. Halsted opened the meeting, and the very first words of his paper 
were of great significance. He said: ''It is especially true of breast 
cancer that the surgeon interested in furnishing the best statistics 
may in perfectly honorable ways provide them/*^ I believe that if 
we are going to attempt a comparison of different methods of operating 
we must have some definite standard which we accept. I know no 
better standard than to take all the cases that come into a general 
hospital under the diagnosis of the disease in question and to record 
them diuing the period under observation, with a statement as to what 
percentage of that number was considered suitable for the attempt at 
radical cin-e, and what results were obtained. I am sorry that the idea 
has not taken hold better in the general reporting of cases. I was 
looking at a report of cases of cancer of the lip recently published, and 
in that report all the cases in which an attempt to trace the patient 
after operation by letter had been unsuccessful, were wiped out of the 
record. It has been our experience in Boston that the patients from 
whom one cannot hear are usually those that have actually left this 
world for another, and I fully believe that patients untraced should be 
coimted as failiures. I believe also, as Dr. Meyer stated so emphati- 

* Halsted, W. S. : The Results of Radical Operations for the Cure of Carci- 
noma of the Breast, Ann. Surg., 1907, xlvi, 1. 



92 , PROCEEDINGS 

cally, that one of the great functions of the social service departments 
of these hospitals should be the tracing of end results. At the Hunt- 
ington Hospital we have an exceUent social service worker who has 
established a standard in regard to these things, and we do not con- 
sider a record closed when the patient has been discharged until we 
have definite information of the end result. We keep them as live 
cases until we know that they are dead. 

3. Primary Spontaneous Tumors op the Ovary in Mice — Studies 

IN THE Incidence and Inheritability op Spontaneous 

Tumors in Mice. Fourteenth Report 

Miss Maud Slye, Miss Harriet F. Holmes, and Dr. H. Gideon, Wells 
(Chicago) : 

SUMICARY 

Among 22,000 mice of the Slye stock d3dng natural deaths at all 
ages were 44 with spontaneous primary ovarian tumors, not including 
simple ovarian cysts. Of these, 38 had simple benign solid papillary 
adenomas, only occasionally with slight cyst formation; 1 showed a 
typical papillary csrstoma, and 1 a t3rpical solid teratoma containing a 
great diversity of tissue elements. A second case of this type has 
since been found in the first 25,000. Of the 38 cases of solid papillary 
adenomas, 19, or 50 per cent were bilateral, so that there were 57 
tumors of this class. There were 4 unquestionable primary malignant 
tumors of the ovary, all showing the '^ mesothelioma" type of growth 
characteristic of malignant tumors derived from the sex glands; one of 
these produced perirenal metastases. One other tumor of the same 
t3rpe was primary in either the ovary or the adrenal. Two round-cell 
sarcomas were found, arising either from the ovary or from some other 
organ, while 2 other sarcomas had produced secondary growths in the 
ovary. Of the 44 mice with primary ovarian tumors, 26 had tumors 
in other parts of the body. 

In the literature were found reports of eight other cases of benign 
tumors arising in the ovaries of mice, all exhibiting the same character* 
istics as the tumors described in this paper. 

discussion 

Dr. F. C. Wood (New York) : Statistics of this sort are of vital impor- 
tance in all experimental work in cancer, since they show that if animals 
are kept to a suflicient old age there is a very large incidence of tumors 
of all varieties. I think that we who experiment with animal tumors 
are still supposed by the clinical fraternity in medicine to be working 
with something which is entirely different from and not in the least 
comparable to human tumors. But as observations on animals are 
being extended over larger series of mice, and to other species, as white 
rats, dogs, guinea-pigs, etc., we are finding instead that tumors abso- 



PROCEEDINGS 93 

lutely comparable in morphology and biological qualities occur in 
many of our domesticated animals. This is interesting^ as furnishing 
an argument against the current opinion that tiunors are a disease of 
civilization, and primarily of the educated classes, and not a disease 
which is widespread and generalized through all groups of society. 
The statistics of the Metropolitan Life Insurance Company, which 
show a greater incidence of cancer among the laboring classes, point 
in the same direction, that is, to the fact that there is no immune 
class. Obviously, in the manual laboring classes the conditions of 
irritation and of infection (for instance, S3rphilis, which, while not 
causing cancer, facilitates its occurrence) result in a larger proportion 
of tumors in that group, and as we study mice we find that larger 
numbers of tiunors spontaneously appear in these animals. I do not 
doubt that the same would be foimd to be true of wild mice if we could 
keep larger numbers to old age. 

Dr. James W. Jobling (New York) : I should like to ask Dr. Wells if 
these ovarian tumors were observed more frequently among the ''tumor 
strains" of mice described by Miss Slye. 

Dr. Wells: In reply to Dr. Jobling's question: Miss Slye had hoped 
to be able to discuss that featiure, as, of course, the study of heredity 
is entirely her part of the work; but on account of illness she was not 
able to get the material together. We have had strains for twenty-five 
or thirty generations producing enormous numbers of mice with no 
tumors. It is perfectly safe to say that mice of certain strains are 
more likely to have tiunors of the ovary than are those of other strains. 
It is a familiar fact that in the strain of mouse tumor that is used in 
most laboratories in America, that derived from the Abbie Lathrop 
stock in Granby, Massachusetts, cancer of the manmiary gland is the 
usual type. Very few cases of tumors of any other tissue have been 
described in this stock; but Haaland's reports indicate that in the 
mice in the Imperial Cancer Research Fund in London, tumors of 
other organs are quite common. I may recall Miss Slye's observa- 
tions, previously reported, that with tiunors of specific organs a very 
definite relationship to heredity is shown. She reported at one time 
twenty-eight cases of primary tumor of the testicle, all, with one 
exception, arising in mice of one definite strain. The exception was 
in a mouse which had been bitten on the testicle, and which belonged 
to a strain in which sarcoma was very common. The most striking 
series which Miss Slye has had was the liver tumor. You may recall 
that the literature of mouse tumors had previously shown but one 
case of primary tumor of the liver; since then one or two others have 
been described. Now Miss Slye has bred, from a mouse with primary 
tumor of the liver, a strain in which there have developed one hundred 
cases of primary growth in this organ. While there are no figures on 
the ovarian tumors, we know that most of them arise in certain strains, 
though whether strictly within the strains, as in the case of the testicle 
and hver tumors, we are not yet prepared to say. But in such experi- 



94 PROCEEDINGS 

ments as these, certain difficulties must be faced. Thus Tyzzer and 
Haaland have described squamous-cell neoplasms of the lung, a t3rpe 
of which we 'have seen only one or two examples. Hence minute 
differences must exist, making one strain Uable to one variety of tiunor 
in any given organ, and another liable to a type slightly different. 

Dr. C. C. LMLe (Cold Spring Harbor, N. Y.) : How would you explain, 
on the basis of a single Mendelian factor, the occurrence oi these clearly 
demarcated strains which show specialized types of tiunors in particular 
organs or groups of organs? 

* 

Dr. Wells: I am not prepared to discuss that aspect of the work at 
all; that I leave entirely to Miss Slye. I do not consider myself com- 
petent to discuss the matter of genetics in these problems. 

Dr, Little: I would like to point out that the result outUned by Dr. 
Wells is very interesting in support of the theory that the hereditary 
nature or susceptibility to the occurrence of spontaneous tumors 
depends on more than one hereditary factor. This is evidenced by the 
clear ability of a particular family to localize its susceptibiUty to tumor 
in a particular organ, and is further evidenced by the abiUty of other 
families to combine tumors of different organs. Such k result is par- 
ticularly characteristic of the action of many factors in heredity, rather 
than of only one. I mention the fact because of the possibiUty that 
too much emphasis may be placed on a single factor as the cause of all 
cancer. The specificity of tissues is so great that it would be surpris- 
ing if such evidence as that given by Dr. Wells were not found when 
so careful a study was made. 



N 



Dr. WiUiam C. Stone (New York): I am very much interested in 
this presentation because of my own study of ovarian tumors in the 
human subject. In the first place, as to diagnosis: Dr. Wells spoke of a 
certain niunber of these tumors as possibly endotheliomata. Similarly 
in the hmnan cases, in going over the literature of the so-called Kru- 
kenberg timiors, one finds numerous instances in which the picture has 
been interpreted by many as an endothelioma, by others as a sarcoma, 
and by still others as a carcinoma. Then, as regards the observation 
of the occurrence of the tiunors in both ovaries, and the difficulty of 
explaining this in human patients, also, in a large number of instances 
the ovarian tiunors are of seconda^ nature, and the primary site is 
elsewhere; but there are numerous cases in which no piimai^ site can 
be found, and the iacidence of the tumor in both ovaries is unexplained. 
There is one diGFerence, however, if I understand Dr. Wells correctly, 
between human and mouse tumors. In the majority of cases the so- 
called Krukenberg tumors in women seem to be secondary in the 
ovaries. If one considers carefully the descriptions of those reported 
as primary, one is imable to exclude the possibility that in the major- 
ity there was a primary site elsewhere. We found it most frequently 



PROCEEDINGS 95 

in the stomach, gall-bladder, appendix, or some other part of the 
intestinal tract.. 

Dr. Wells: Of course, in these cases the histology is not that of the 
typical Krukenberg tmnor. The tumors are bilateral, but they do 
not resemble the Krukenberg timior. They do show characteristics 
which I am quite sure would be cause for a diagnosis of endothelioma 
or of sarcoma; in fact, before I had studied these timiors carefully I 
labelled a good many of them endotheliomata of the ovary. It may be 
interesting to know that every once in a while we see in bovines bi- 
lateral tumors of the ovary; some of those I have seen from the stock- 
yards in Chicago are histologically similar to the Krukenberg timior 
although nothing is foimd to indicate a primary tumor elsewhere in 
the cattle. In mice there have been reported only five or six cases of 
primary abdominal tmnors which could give rise to Krukenberg 
timiors; so that here we have timiors which are bilateral, and which are 
distinctly not secondary tumors. Why they should be bilateral in 50 
per cent of the cases, and why bilateral tumors are so much more likely 
to occur in the ovaries I am unable to explain. 

4. The RdLE or Neoplasia in Parasitic Diseases op Plants. 
Dr. Isaac Levin and Dr. Michael Levins (New York) : 

Previous investigation by the authors on the crown gall have demon- 
strated that while this condition frequently acts in a manner analogous 
to animal cancer, the cellular proliferation is primarily a reaction to 
the invasion of Bacterium tumefadens. The previous experiments 
were conducted on annuals, biennials, or deciduous trees, in which the 
period of growth of the host as well as of the crown gall is normally 
interrupted. In these experiments, some of the galls are benign to 
the host and behave in a manner more analogous to a scar, a cheloid, 
than to cancer. In a comparatively small percentage of cases the galls 
act as true malignant tumors. The parts of the inoculated stem 
become necrotic above and even below the point of inoculation. 

The present investigation was conducted on the rubber tree (Ficus 
elastica), which is an evergreen perennial plant and grows indoors, so 
that the crown galls may be watched for long periods of time without 
interference by secondary contaminations. These experiments have 
shown that twelve months and more after inoculation, nearly every 
crown gall produces a necrosis of the inoculated branch, though at 
first the gall may attain a large size without apparent injury to the 
host-branch. Then frequently, without any additional increase in 
the size of the gall, the branch becomes necrotic both below and above 
the gall. This necrosis increases in a centrifugal direction from the 
gall; thus the tip of the branch may still be alive while a part of the 
branch below the gall is necrotic. 

These phenomena cannot be due to impairment of nutrition since 
the crown gall frequently attains its largest size without a concomitant 



96 PROCEEDINGS 

necrosis, and the latter is always subsequent; nor can the necrosis be 
due to the action of some toxic substances produced by the crown-gall 
cells. It is difficult to conceive of such a rapid change in the metabolic 
chemical functions of a cell. The most plausible explanation is a 
change in the reactivity of the host tissue. 

For a time the Bacterium tumefaciens produces a progressive reaction 
in the injured and surrounding tissues of the host, which manifests 
itself in proliferation of cells and formation of a crown. Sooner or 
later this is followed by regressive reaction which manifests itself in 
progressive necrosis. Since plants lack the lymphoid tissues and can- 
not react to parasitic invasion by inflammation, the above described 
methods of reaction are the only possible ones. 

Thus, the neoplasia in the crown-gall disease is primarily a protective 
reaction of the host tissue to the invasion by Bacterium tumefaciens. 
The morphological studies by the writers of two other parasitic dis- 
eases, clubroot of cabbage (PUismodiophora hraesiccR), and potato-wart 
disease, or potato cancer (caused by Chryaophlyctis endcbiotica) show 
that identical reactive cell proliferations and formation of new growths 
may be caused by other parasites besides Bacterium tumefaciens. 

It is possible that in most parasitic diseases of plants both neoplasia 
and necrosis take place, thoi^h the former may be so insignificant and 
transitory that it evades detection. 

DISCUSSION 

• 

Dr. Wells: It is pleasant to hear that potatoes and cabbages contain 
items of interest besides calories and vitamines. This subject is, of 
course, of great importance, because the extremely interesting work of 
Dr. Smith has attracted so much attention, and there has been a lively 
discussion as to whether these processes in plants are really to be con- 
sidered as true neoplasms. I wish to express my personal apprecia- 
tion of this contribution because it must be determined whether these 
things in plants are true tumors or not, since their interpretation will 
have much bearing on our evaluation of the work in cancer research. 

5. A Phase op Tumor Biology 
Dr. Frederick Prime (New York) : 

SUMMARY 

In the past few years much confusion has arisen on account of the 
various results reached by different cancer investigators doing prac- 
tically the same line of research. A great deal of this confusion is 
probably due to the use of tumors whose biological characteristics are 
really imknown to the investigator. AppUcations are made to this 
laboratory every year by workers on the cancer problem who desire a 
carcinoma or sarcoma which they wish to propagate for a short time in 



PROCEEDINGS 97 

order to carry out some investigation. They then report their results 
on a few dozen animals, whereas if the characteristic action of the 
tumor over numerous generations had been known to them, or if they 
had used a larger number of animals their results would have been 
very different. For instance, the Jensen rat sarcoma is reported in 
one paper to take in almost 100 per cent of animals inoculated. In 
running back over our books for the last six years we find that in 5400 
ftTiinriRla surviving at the end of three weeks we have only one series 
in which there was 100 per cent of takes, and only a few in which the 
takes were as high as 90 per cent. If our experiment had been done 
in March, 1914, for instance, we should have had 100 per cent takes 
in our controls, whereas if it had been done in July, 1916, we should 
have had only 20 per cent positives. In another instance important 
conclusions were drawn from two series of tumor 63, one with a high 
percentage of takes, the other with no takes, but our books show that 
the same fluctuation has occurred in routine transplantation. The 
results vary from month to month and from year to year, but if the 
averages for the months are jaade the variations are less marked. Of 
all our tumors, the mouse sarcoma Crocker Fund No. 180 gives the 
highest percentage of takes and is the most consistent, the number of 
takes never having fallen below 85 per cent. The Ehrlich mouse 
sarcoma is a close second to this and in only one month did this fall 
below 80 per cent of takes. 

Certain members of our staff, I myself among them, have had the 
general impression that our tumors grfew less well during the summer 
months, but upon analysis this proved to be quite incorrect; the deati^ 
rate among the animals may have been higher, but the tumor growth 
rate did not show any such change. For six years the Jensen rat 
sarcoma had its highest number of takes in August, and its lowest in 
July. The Flexner rat carcinoma had the highest nimiber of takes in 
April, and the lowest in November. It behooves us, therefore, to 
know and study our tumors carefully before starting any investigations 
on them, and to recognize the fallacy of using small numbers of animals 
in arriving at a conclusion. 

DISCUSSION 

Dr. 8. R. Benedict (New York): I think it would be interesting if 
Dr. Prime could give some more definite statements as to how much 
these fluctuations would influence conclusions, that is, whether t|^e 
results would be diametrically opposed through an increased number of 
animals, and whether he would recommend that an experiment be 
based on 5000, or on 500, or on 50 animals. A general criticism of this 
t3rpe calling attention to the number of animals might be misinter- 
preted unless it is made more specific. In the matter of transplanting 
a given tumor, for instance, we have quite exact data, and if we trans- 
plant from an experimental animal, using the same tmnor, we should 
get a similar growth in both series, except for the experimentfd factor. 



98 PROCEEDINGS 

I should like to ask how large a series would have to be, and whether 
it must be repeated for several months before definite conclusions can 
be drawn. 

Dr. Prime: I think the criticism is exceedingly pertinent. We feel 
that a good deal of confusion arises with men working along the same 
line who report such varying results from exactly the same tumor, 
because they do not quite know the biology of the tumors. One may 
report results with a certain series or a certain tumor and another give 
data of an opposite nature, but that does not necessarily mean that 
one or the other is incorrect. If a little more were known about that 
tumor, it woidd probably be found that neither was making very great 
errors. I think one advantage of a laboratory of this kind is that we 
have tumors in such large series, and that our records are always avail- 
able, so that anyone can refer to a timior with which he is working. 
But so many of the people who come here for timiors do not care what 
they get or what happens to it. They want a tmnor they can grow 
in one or two generations. I do not' think conclusions should be drawn 
from experiments with 25 or 50 animals.' I think each series should 
contain one or two himdred animals, and that one series alone should 
not be taken as a basis for conclusions^ As you see here our results 
vary from month to month; and if experiments were made on several 
series and the results averaged, I think they would be very much more 
valuable than are deductions drawn from one or two small groups, as 
is often done. It is simply that we want to avoid the error of making 
too sweeping conclusions from too few animals and from a lack of 
knowledge of the growth ratfes of these timiors. I think that eventually 
if experiments are &ade in this country with tumors whose growth and 
behavior is known, and with animals whose behavior is known, results 
from the various laboratories would be very much more consistent. 

Dr. Wood: The question has other phases, also, such as the statistical 
study of tumor regression after taking and the study of the tumor 
growth rate. In some instances the number of takes is most impor* 
tant, for instance, when attempting to establish immunity against a 
tumor it is vital that we know that the tumor is not one which estab- 
lishes immunity against itself. Some workers have reported experi- 
ments in which they claimed to have established immxmity by means 
of treatment of the animals, but they overlooked the fact that the 
tumor used was one which inmiunizeo against itself. In experiments 
in which attempts are made to influence the growth of a tumor by 
therapeutic means, it is necessary to use a tumor in which the number 
of spontaneous disappearances is negUgible. Our no. 180 is such a 
tumor, and because of this fact and because of its easy inoculability, 
we usually supply it to those who wish to make therapeutic tests. One 
naive experimenter, after working with this tumor, wrote that he 
wished we would send him a better one, because although the tmnor 



PROCEEDINGS 99 

he had previously used was curable by his treatment, this no. 180 was 
worthless because it was too resistant. The value of his cancer cure is 
evident. 

As to the number of animals to be employed: we are in the habit of 
using small series of animals over a long period of time; instead, for 
instance, of using three himdred animals for an experiment, and one to 
three himdred as controls we use two sets of twenty-five ftnimfLla each, 
and then repeat the experiment several times, employing a variety of 
types of tiunor, both carcinoma and sarcoma. We thus get long 
series which can be compared with each other. A few months ago 
there was published a statement concerning certain effects of x-ray, the 
conclusions being based on the fact that in twenty or twenty-five ani- 
mals, treated in one way, there were no tiunor takes, while in the 
control animals there were 80 to 90 per cent of takes. It was, there- 
fore, assiuned that an immimity was proved. All it was necessary to 
do to show that the point was not proved was to open one of our record 
books of the tumor this experunenter was using and find a page on 
which were listed twenty-four animals which had been inoculated with 
the same tumor on the same day, without a single take, while the next 
lot of twenty-four under the same conditions showed 60 to 70 per 
cent of takes. The only way to avoid these errors is to repeat the 
experiment time after time, not necessarily with a large number of 
animals, and in this way eliminate random fluctuations of which we 
know nothing, but which are due to the fact, too often forgotten, that 
an animal is not a test-tube, and that the tumor graft, also, varies in 
its biological activities. 

As to the question whether or not we can stimulate or retard the 
growth rate of a tumor, the evidence which we have accumulated 
proves that it is impossible to draw any accurate conclusions as to 
rate of growth. Some 3000 grafts of tumor no. 180 were planted in 
healthy mice from one dealer, allowed to grow for three weeks, and then 
excised and weighed. It was assumed that the averi^e weight of 
these tumors would furnish a useful standard, yet in the next series of 
200 tumors the average weight was foimd to be double that of the 
previous 3000. When precautions are taken to have the site of inocu- 
lation, the strain, the weight and the age of the animals, and the 
method of feeding the same in all cases, and yet great fluctuations 
occur, we are justified in saying that no conclusions as to growth rate 
can be drawn. As the result of experiments in which only a small 
number of animals was used and no controls were kept, some obser- 
vers have claimed to have stimulated tumor growth by chemicals; 
but while such claims may be correct, they are absolutely unproved 
because there are no available measurements by which growth rate 
can be determined. 

These are some of the compUcations of this tumor work, and the 
only way to eliminate them is to know the biological qualities of the 
tumor we are using, and to adjust the numbers of animals used to the 
conditions of the experiment. 



100 PROCEEDINGS 

Dr. Stone: It seems to me that the purpose of this presentation is an 
exceedingly good one. It brings to my mind the experience we are 
repeatedly having at the Memorial Hospital in regard to radium. 
Men go to Pittsburgh, for example, and make a contract for radium. 
Then they come to the Memorial Hospital to learn how to use it; and 
after a few days they return to Pittsburgh and get their radiimi. The 
conclusions drawn from their work certainly can be no more accurate 
than those which are drawn from the observation of the effects of 
therapeutic measures on a few animal tumors. 

6. The Influence of Certain Diets upon Tumor SuscEPriBiLnT 

AND Growth in Albino Rats 

Dr. K. Sugiura and Dr, Stanley R. Benedict (New York) : 

SUMliARY 

Curves constructed so as to show the linear €md the percentage 
growth of the fetus and of the Flexner-Jobling rat carcinoma in nor- 
mally fed animals indicate that there is quite close agreement in growth 
between the two (the carcinoma and the fetus). 

The experiments reported were designed to determine whether 
general differences in diet influence tumor growth independent of 
special deficiencies, and to study the effect of certain specific defi- 
ciencies in diet upon tumor susceptibility and growth. A complete 
diet, composed of banana, 83 per cent, yeast, 0.5 per cent, and protein- 
free milk, 0.5 per cent, was foimd to yield the same percentage of suc- 
cessful inoculations and the same rate of growth as was secured with a 
diet of wheat bread and whole milk. The elimination of certain 
accessory factors in the banana diet was found not to influence tumor 
growth except when the diet was restricted wholly to bananas. In 
this latter case, the percentage of successful inoculations was the same 
as in control animals, but the rate of growth of the tumor was markedly 
retarded. Such dwarfed timiors showed no histological changes from 
the normal, and resumed normal growth when the animal was placed 
upon a normal diet. 

discussion 

Dr. Wells: Dr. Benedict's paper giving the curves of fetus and tumor 
growths recalls a discussion which I heard when a paper from Ehrlich's 
laboratory was presented. Some German, with a mathematical turn 
of mind had figured the rate at which tumor cells grow. At the end 
of a certain niunber of days, he said, one would have enough tumor to 
inoculate ten more animals, and at the end of three years the total 
tumor tissue would have formed a cube so large that it would take a 
ray of light 105 years to pass one side of it. Some of the biologists 
present said there was nothing remarkable in that, because fetal tissues 
had been demonstrated to have a fully equal capacity for growth. 



PROCEEDINGS 101 

Dr, Wood: I did not know when I spoke of tumor growth rates that 
Dr. Benedict was going to present an3rthing which turned upon the 
details of growth rate, and I still hold the same views that I expressed 
a few minutes ago. My opinion of Dr. Benedict's experiments is that 
they may be ri^t and they may be wrong. The only way to prove 
them right is to take the two significant series and repeat them on a 
group of twenty-five rats, some ten times. This will eliminate random 
fluctuations. Then if the averages show a striking difference between 
the dieted and the control groups the fact may be accepted that diet 
influences the t3rpe of tumor used, though not any other tumor, and 
especially not a primary timior. I can take from my own record 
books pages that show very much diminished tumor growth rates 
running over a period of two or three months, followed by a period of 
very abundant growth, although the animals were untreated. I 
cannot say, therefore, that Dr. Benedict is right or is wrong and the 
only way to get proof either way is to repeat the experiment under a 
great variety of conditions. It is the same problem as the question 
whether radium does or does not stimulate a tumor. Many workers 
believe that threshold doses of radimn do stimulate but I have been 
unable to prove it. 

7. The Effect of Combined Heat and Radiation upon 

Transplanted Animal Txtmobs 

Dr, George L. Rohdehburg and Dr. Frederick Prime (New York) : 

SUMMARY 

The investigations of Loeb, Stevenson, and others have shown that 
neoplastic cells are killed in relatively short periods by comparatively 
low degrees of heat. In the present experiments mice and rat tumors 
were cut into small fragments suitable for inoculation, exposed to 
varying degrees of heat for varying periods while in Ringer's solution 
in a water-bath, and then inoculated into animals, the lethal effect of 
the manipulation being indicated by the percentages of takes in the 
inoculated animals. In this fashion exposures to 40, 41, 42, 43, 44, 45, 
and 46^0. were made for 15, 45, 75, 135, and 195 minutes. The lethal 
effect of heat was manifested first with 40^C. at the end of 195 min- 
utes, and with 41®C. at the end of 75 minutes; with 44® there were only 
35 per cent of takes at the end of 45 minutes. 

Bovie has shown that albumins which have been radiated coagulate 
at much lower temperatures than those not radiated, and it was thought 
it might be possible to apply this principle to the killing of the cancer 
ceU. Fragments of tumor prepared as were the fragments used in the 
heat experiments were exposed to rr-ray for periods varying from 
10 to 30 minutes, 30 minutes representing three er3rthema doses. 
Directly after radiation they were exposed to var3dng degrees of heat 
after the method outlined in the previous paragraph and then inocu- 



102 PROCEEDINGS 

lated into animals: It was found that whereas heat alone at a given 
temperature was not lethal, and x-ray in a given dosage was also not 
lethal, the two when combined were lethal. 

In another series of experiments the process was reversed, heat being 
applied first, and radiation afterwards. The same effect was noted, 
the sequence being apparently immaterial. 

These results suggest a new method for treatment in human cases, 
and experiments are now under way in which a practical application of 
the principle is made on both human and spontaneous tumor animal 
material. 

DISCUSSION 

Dr. Greenough: We have felt very strongly that the observation of 
Dr. Bovie in regard to heat sensitization was one of possibly very great 
clinical importance, but that further work should be done idong experi- 
mental Hues before any serious attempt is made to apply it to Uving 
patients. The point of the greatest importance is that a degree of 
heat which under ordinary circmnstances can be withstood perfectly 
well by the individual cell results in the death of that cell if it is applied 
after radiation. The temperatures which Dr. Rohdenburg has indi- 
cated on the Qharts apparently are such that a definite effect can be 
obtained in a certain small percentage of cases from the temperature 
alone; to that extent, therefore, the experiment is not quite comparable 
to Dr. Bovie's original observation; but I am very much interested in 
the results, and I feel that without question this opens up a field for 
clinical application in the actual treatment of cases. 

Dr. Wood: I think everyone realizes now that we have reached a 
point where we can say definitely whether or not radium and x-ray will 
cure any malignant tumor. The only question is will the patient sur- 
vive the dose. In other words, we have come to a point where it is a 
question of operative mortality as compared to x-ray or radium mor- 
tality. It is perfectly possible to obtain the absolute destruction of 
cancer cells by a sufficient quantity of x-ray; and the same thing can be 
done with radimn. The only question is the practical application. 
In my opinion, at the present time most internal tumors require so 
much radiation that serious, very often fatal, damage to the normal 
organs would inevitably be inflicted. For example, the lethal dose to a 
carcinoma cell of a tumor 10 cm. below the skin would be between 
sixteen to twenty erythema doses, if a highly filtered x-ray were used. 
I have seen quite serious general disturbances to the intestinal tract 
result from two erythema doses. The recent German literature, also, 
contains reports of a number of deaths, some apparently from intes- 
tinal obstructioQ, following heavy doses of x-ray. The problem is to 
kill every cancer cell, although it is recognized that much benefit can 
be obtained by the destruction of the main portion of the tiunor, the 
central areas, for example, which are poorly vascularized and easily 



PROCEEDINGS 103 

subject to necrosis from thrombosis of the capillaries of the tumor. 
But the destruction of the peripheral portions of the tumor, which are 
well vascularized, is quite another matter. Hence, a very large dosage 
is required to kill absolutely all the tumor cells, either in the test-tube 
or in the animal itself. Tlius, although a point has now been reached 
where we can say definitely that cancer can be cured by radiation, 
with the dosfitge determined experimentally in this laboratory, the 
question still is how we can also save the patient's life. With this 
point kept clearly in mind Dr. Rohdenburg started his experiments. 
The original experiments Dr. Prime and I began some four or five 
years ago. The lethal death-points for cells were determined and are 
standard, so there can be no difficulty in reproducing these results 
anywhere if tumor 180 is used. An x^ray machine can be caUbrated 
biologically. My clinical experience shows that in a rapidly growing 
carcinoma in man the death point is about the same as in 180. The 
slow growing squamous-cell epitheUomata seem to have extra resis- 
tance, owing to the abiUty of the cells to comify and practically to 
enter a resting stage. It is now accepted that the lymphosarcoma and 
many of the bone tiunors are more susceptible to x-ray than are the 
very rapidly growing highly malignant carcinoma or sarcoma. The 
maximimi practical dose of x-ray or radium which the patient can 
survive remains to be determined, but if it is remembered that there 
is, after all, a surgical mortality and that imtreated cancer is fatal, we 
are justified in giving radiation pretty close to the limits. Most 
radiation treatments are far beneath the killing dose for the cells of 
most of the internal tumors. I do not believe that the induced con- 
nective tissue reaction around the tmnor cells kills them. It may 
encapsulate them for a while, but usually the ceU enclosed with scar 
tissue finally begins to grow and destroys the patient. After having 
given the maximum dose of x-ray which the skin of the patient can 
stand, we might be able by thermo-electrical devices to apply heat 
which does not seriously damage the skin and which produces death 
of the cells. The morphology of cells killed by heat is practically the 
same as that of cells killed by x-ray or radium, and I think that this 
paper is important from a therapeutic point of view, as well as a scien- 
tific one. It may be possible by some high frequency method to add 
to the semi-lethal dose of x-rays a semi-lethal dose of heat without 
causing the death of the patient and that, I think, these experiments 
clearly show. The practical application will require long careful study 
on human beings. 

Dr. William Duane (Boston) : In regard to the practical application 
of heat and x-rays, this was tried very extensively in Paris seven years 
ago, and there was claimed a good deal for the combination. There 
was one difference between that procedure and the one reported here, 
the difference being that the two destructive agents were applied at the 
same time. Keating-Hart warmed the himian tumor tissues and at 
the same time applied x-rays, and I imderstand that here heat was 



104 PROCEEDINGS 

applied either before or after radiation. It would seem probable that 
the application of the two at one and the same time would produce a 
greater eflfect. As regards the ability to destroy any tumor by radia- 
tion, if an intense enough radiation is used almost any organic com- 
pound can be destroyed. Water can be decomposed by the action of 
the rays. 

Dr. Rohdenburg: I would like to say that these experiments differ 
from those previously reported in that lower degrees of heat were used 
over a much longer period. In the recent German literature the com- 
bination is reported to have been tried, but tried by combining diath- 
ermy produced by seven amperes of current for a period of thirty 
seconds. This dosage produces coagulation necrosis and a typical 
bum. Our aim is to obviate this factor of bum. The tmnors we have 
treated with high frequency, using 25 milliamperes per square inch 
of electrode surface and continuing the application for twenty minutes, 
show scarcely any recognizable change in the tissue for a period of a 
week or ten days. This dosage produces a temperature between the 
electrodes of 40®C., or 104.8**F. which is very readily withstood by any 
normal tissue, and many degrees below the temperatures produced by 
Keating-Hart or the more recent investigators. 

8. Intbestitial Injection's op an Active Deposit op Radium 

Emanation in a Rat Carcinoma 

Dr. Hahey J. Bagg (New York) : 

summabt 

In this experiment a definite attempt was nxade to use an '^ active 
deposit^' of radiimi emanation as a local agent, employed in the form 
of a solution, to control the growth of an experimental rat tumor — 
the Flexner-Jobling rat carcinoma. The writer gave a preliminary 
report of the results of treating twenty-one animal tumors. 
^ Interstitial injections of a radio-active salt solution of radium emana- 
tion were found to retard materially the growth of the tumors and in 
some cases to cause their regression, resulting in characteristic cellular 
changes in the timior tissue. Definite characteristic, histological 
radium changes were noted — ^frequent hyperchromatic and homog- 
eneous nuclei, cellular hydrops, fairly extensive central necrosis, and a 
terminal onset of fibrosis. 

It was found that comparatively large doses of radium could be in- 
jected into rat tumors without the escape of the radio-active solution 
into the siu'rounding tissues. (A comparatively small intravenous or 
subcutaneous injection of the same solution invariably results in severe 
reactions.) The rapid decay of the radium when used in this form, in 
addition to the walling-off effect of the timior capsule, appears to be 
sufficient to confine the radium action to a desired zone of tissue. This 
was an encouraging observation from the standpoint of human therapy. 



PROCEBDINGS 105 

DISCUSSION 

Dr. Duane: I think it would be well to emphasize the very great 
danger of this method of using radium or radio-active substance. The 
alpha rays are utilized, and since of the total activity of all the beta, 
ganoma, and alpha rays, about 90 per cent is due to the alpha radiation, 
it is evident that the alpha rays produce relatively tremendous effects. 
It is well to bear in mind, therefore, that this method of treating tumors 
should be employed with very great care. Over 100 millicuries of 
deposited activity would be a dangerous dose for any himian being. 

Dr. Benedict: I should Uke to ask if this is not simply a question of 
the size of the tmnor mass; that is, one might destroy a small tumor, 
but affect only the center of a large one, making it necrotic. 

Dr. Stone: It seems to me that both of these methods, the intra^ 
venous injection of the active deposit and the local infiltration of the 
tumor, represent a type of experimental work which is most desirable, 
but, as Dr. Duane has just said, we must use extreme caution in apply- 
ing this practically. The general constitutional effects from intra- 
venous injections are very marked. So far we do not know the con- 
stitutional effects from the infiltration method, because it has not been 
applied except in two or three instances, and then very cautiously. 
It does seem, however, as if we might make use of the alpha ra3r8 in 
this infiltration method, which, as Dr. Duane has said, comprises so 
much of the energy of radium, and which apparently we cannot use 
in any other way. ^ There is another point in regard to the practical 
application of this infiltration method. PersonaUy I dislike to thiiik 
of radiimi used either by this method or by the surface application to 
the extent that we get actual destruction or caustic effects. I conceive 
of using radiation for a strictly biological effect, without causing abso- 
lutely local destruction, for years prior to the use of radio-activity aJl 
of us had experience with the use of the cautery, and we knew just how 
much tissue we could destroy with. that. With the use of radio-activ- 
ity we do not know just how extensive the destruction is, how long it is 
gjoing to continue, or what the destructive effects upon the normal 
tissues in the neighborhood of the timior are. The possibilities of 
destruction are so great that we must use extreme caution. In making 
a practical application of this infiltration method I think we must 
begin with the very smallest dose so as to get the so-called biological 
effect, and not a local caustic effect on the timior. itself. 

Dr. WiUiam B. Coley (New York): The work of Dr. Bagg is ex- 
tremely interesting and valuable, but I believe that the cautions given 
by Dr. Stone and Dr. Duane are timely. At the Memorial Hospital 
Dt. Janeway has used the method in a considerable number of cases in 
human beings, beginning with small doses and increasing to fairly good 
sized ones; but so far none of the results gives us reason to believe that 



106 PROCEEDINGS 

it is likely to be of permanent value as a therapeutic measure. I 
believe, as Dr. Bagg points out, that the cases in which good results 
are obtained are those in which the tumor is localized by the surround- 
ing tissues, and if that is so, why can not one use the bare tubes which 
will remain localized, and not get into the tissues and damage the 
kidney, Uver, etc.? I believe with Dr. Wood that we can always give 
enough radium to kill the cancer cells, but not in most deep seated 
cancers without killing or injuring the patient. 

Dr. Bagg: Dr. Benedict has brought up the subject of the relation of 
tumor size to the subsequent reaction to the treatment. I have found 
that in most cases the central portions of the treated tumors were 
necrotic, and Dr. Ewing, who has seen my sections, has called my 
attention to the peculiar manner in which the radio-active solutions 
have diffused to different parts of the tumor, causing more extensive 
reactions in certain areas of the tumor than in others. I have tried in 
some recent work to infiltrate the tiunors from within, by first placing 
the hypodermic needle in the center of the mass and then moving it to 
different parts of the periphery of the tiunor, and injecting a small 
amount of the solution in each place. In regard to the apparent 
increase in size of some of the treated tumors, I would say that this is 
' no doubt due to the production of an edematous condition and the 
accmnulation of fluids in the center of the tumor, in which case the 
ring of tmnor tissue was crowded to the periphery, while the actual bulk 
of the tumor tissue was no greater than the amount present at the 
beginning of the experiment. 

9. Factors Underlying Susceptibility to a Transplantable 

Tumor in Mice 

Dr. C. C. LMe: 

SUMMARY 

A sarcoma, J. W. B., which originated in an inbred race of Japanese 
waltzing mice grows upon transplantation in 100 per cent of the ani- 
mals of that race inoculated. It falls to grow progressively in more 
than 99.5 per cent of the common non-waltzing stock mice inoculated. 

Beginning two weeks after inoculation each animal is observed at 
weekly intervals, and the presence or absence of growth is noted. 
When a growth is present the animal is recorded as + ; when no growth 
is found, it is recorded as — . Two groups of animals have been 
observed. They are (1) ordinary non-waltzing stock mice series (N.); 
(2) back-cross hybrids (B. C.) produced by crossing an Fl generation 
hybrid between Japanese waltzing and common non-waltzing mice 
back with the common non-waltzing parent race. 

The present series of experiments deals with growth of the tumor 
from the second to the sixth week after inoculation, inclusive. Many 



PROCEEDINGS 



107 



of the animals showing a growth at the sixth week observation showed 
eventually regression and disappearance of the tumor. The factors 
studied are, therefore, those allowing the initial six weeks' growth of 
the tuinor. This growth may or may not be continued later, depend- 
ing upon the hereditary constitution of the mice used. The animals 
inoculated were divided by age into ten groups, respectively, 2, 4, 6, 
8, 10, 12, 14, 16, 18, and 20 or more days old at inoculation. For each 
of these groups the percentage showing growth at 2, 3, 4, 5, and 6 
weeks after inoculation was determined. The resultis of all groups 
according to weeks after inoculation were as follows: 



B. Cm rmn obmt -f 



2 
3 
4 
5 
6 




27.1 
21.0 
12.4 
13.6 
13.1 



It will be noted that the percentage of animals showing growth in 
the N. series steadily decreases, while in the B. C. series, there is an 
initial decrease followed by a rise to a level at the fourth week at or 
near 13 per cent. The total per cent for the N. series is 11.12:^46 
and for the B. C. series is 17.54^^83. The difference is 6.42=*= 95 and 
is, therefore, significant, being 6.7 times its probable error. 

Studied by age groups with the sexes combined, there is seen to be a 
distinct difference between the two groups. 



▲OB AT ZNOCULATZON 


N.,PSBCBMT + 


B. C. PBB OBMT + 


2 to 10 
12 to 20+ 


12.87db0.6 
9.45=1:0.6 


13.77=fcl.06 
21.58=tl.28 



While in the N. series the lower age group shows the most growth, 
exactly the opposite is true of the B. C. series in which the higher age 
group shows a significantly higher percentage of growths. 

When the sexes are studied separately the males of both series and 
of both age groups are apparently not significantly different from each 
other. 1%e females, on the other hand, show a significantly lawtr 
percentage of growths in the higher age group of the N. series and a 
significantly higtier percentage of growths in the higher age group of the 
B. C. series. The higher age group of both sexes is more mature and 
further differentiated than is the lower age group. Within the higher 
age groups many of the females become sexuaUy mature during the 
period of observation. This gives to them an additional chance for 
differentiation and the assumption of sex limited or other biological 



108 



PROCEEDINOS 



characters dependent upon their hereditary make-up and upon the 
degree of differentiation which their tissues have reached. 



AGE AT 


N. 


B.C. 


INOCULATIOlir 


d* 


9 


e?" 


9 


daya 

2 to 10 
12 to 20+ 


15.70d:1.64 
10.30d=1.20 


19.46=bl.31 
9.39=b0.87 


14.51d=1.51 
15.38dbl.55 


12.12=bl.76 
25.74±2.07 



What we are observing, therefore, is a racial difference. In the N. 
series a steady decrease in percentage of growths, especiaUy marked in 
the females, is found. In the B. C. series there is at first a decrease 
followed by a rise, a level of about 13 per cent growth being reached. 
This level is due for the most part to the presence in the B. C. genera- 
tion of animals which will show permanent progressive growth of the 
tumor. The results of both series coincide with and amplify the 
explanation of the hereditary nature of susceptibility to the J. W. B. 
tumor, advanced by Tyzzer and the writer in 1916. 



A STUDY OF A LIPOMYXOSARCOMA WITH COM- 
MENTS UPON THE ORIGIN OF THE FAT CELL 

VICTOR C. JACOBSON 

Frirm the Pathological Laboratory of the Peter Bent Brigham Hospital and the 
Department of Pathology , Harvard Medical ScJiool, Boston 

Received for publication April 19, 1021 

Neoplasms of mesenchymal origin are frequently the cause 
df much controversy owing to a lack of diJOferentiation into 
tissue which would permit their being placed in a definite 
organoid category. The term ^^liposarcoma" has been used 
to denote malignant change in the connective tissue of a lipoma, 
itself one of the most benign of tiunors, and also to accent the 
fact that the growing fat cells have invaded the surrounding 
tissues or metastasized. A tumor wfaidi presents th« accepted 
criteria of malignancy, derived from fat cells which are atypical 
in size, shape, and staining characteristics, with many mitotic 
figures and invasion of septa and capsule, is of the greatest 
rarity, and a sfaudy of such a growth may throw light upon the 
origin of the fat cell, regarding which there is much conflicting 
opinion. The writer has lately had the opportunity to examine 
a tumor of that type, removed from a patient in the Surgical 
Out-Door Department of the Peter Bent Brigham Hospital, 
and with the kind permission of Dr. Harvey Gushing a brief 
summary of the case is given. 

The patient, a Russian Jew, came to the hospital on November 18, 
1920, complaining of a growth in his right tUgh. His past history 
is unimportant except in so far as it concerns the tmnor. In 1917 
he noticed a firm swelling the size of a walnut beneath the skin on the 
mesial surface of the right thigh. It caused no pain. A few months 
later it was noticeably larger, and massage was instituted for a short 
time. It continued to grow slowly and became a little painful, particu- 
larly in cold weather. He was, however, able to work in his store, 

and he walked into the hospital. 

109 

TSn JOUBNAL OF OAMCBB SBSSAIIOH, VOL. YX, NO. 2 



110 VICTOB C. JACOBSON 

Physical examination was negative except for the local findings. 
On the inner surface of the upper third of the right thigh, beneath tiie 
skin and extending deep into the muscles, was felt a somewhat movable, 
irregularly rounded growth, about 18 x 6 cm. in size, and not tender. 
There were no palpable nodes. On November 19, under novocaine 
anesthesia, the growth was removed by Dr. A. H. Brewster. It seemed 
to have three distinct capsules, but there was no evidence of invasion 
of the surrounding muscles. The outer two capsules were thick and 
fibrous, and stripped away readily from the more delicate layer covering 
the tumor. Most of the blood-vessels entered through a sort of pedicle 
on the deeper portion of the tumor. The dead space was dos^ with 
catgut sutures and the skin with silk. On December 21 the wound 
had entirely healed. On February 3, 1921, the wound was in good 
condition, with no evidence of local recurrence or metastases, and 
the patient felt perfectly well. 

PaOidlogieal examination. The specimen consists of an iiregular mass 
measuring 17 x 8 x 5 cm., and weighing 510 grams. It is surrounded 
by a delicate transparent capsule beneath which ramify a moderate 
number of dilated blood-vessels. The larger vessels converge upon a 
narrow xone on one side which probably represents the pedicle where 
the vessels were clamped off. It is uniformly firm, more so in slightly 
depressed areas between the bulging large lobules into which it seems 
to be divided. A longitudinal incision shows it to consist of two main 
types of tissue. About three-fourths of the tumor is pale, rather pink, 
and opaque, with the consistence of a very cellular tumor. The mass 
is divided into more or less definite lobules by fibrous septa of varying 
thickness. A broad fibrous band apparently invaded by streaks of 
tumor separates it into two main parts, which are in turn subdivided 
into smaller lobules by narrow vascular septa. More or less well 
circumscribed areas suggesting mucoid d^^eration are present in the 
middle portion, irregular wells of mucilaginous fluid bathing the surface 
like honey in a comb. The fluid gives the characteristic reactions of 
true mucin. 

Microscopical examinoHon. Tissue was fixed in 10 per cent neutral 
formalin and in Zenker's fluid. Frozen sections were stained with 
Scharlach R and hematoxylin, other preparations with methylene blue 
and eosin, phosphotimgstic acid-hematoxylin, aniline blue-acid fuchsin, 
osmic acid, Mayer's mucicarmine and mucihematein, and Verhoeff's 
elastic tissue stain. A study of the sections gives the following facts. 

The tumor is surrounded by a thin compact layer of connective tis- 
sue in which are a moderate number of elastic fibers. The capsule 



STUDY OF A LIPOMYXOSARCOlfA 111 

sends prolongations into the tumor dividing it into smaUer units. 
Elastic fibers and thin-walled arteries and veins accompany the septa 
and many fine capillaries are present in the stroma^ which consists of 
a deUcate collagenous reticulum in which the tumor cells lie, and which 
appears to arise for the most part from the blood-vessels. The tumor 
is very cellular except in small areas which have undergone mucoid 
change, where the tissue suggests that of the umbilipal cord. In shape 
the cells vary from medium sized spindle cells through intermediate 
forms to the more or less roimd outline of mature fat cells. The nucleus 
follows the general shape of the cell and has a densely staining nucleolus. 
Practically all the ceUs contain refractive fat droplets which stain with 
Scharlach R and reduce osmic aeid. In the spindle cells the droplets 
are, as a rule small, almost alwa3rs smaller than the nucleus, and tend 
to localize in the vicinity of the nucleus. These spindle cells often 
have delicate wavy processes of some length. In the intermediate, 
polyhedral forms the fat drops become lai^er and coalesce, the nucleus 
stains more deeply, and mitotic figures are seen in large numbers. The 
dividing cells have a large fat content. The more mature roimd cells 
are distended with a single droplet of fat which occupies practically the 
entire cytoplasmic space and flattens the nucleus, giving the typical 
signetr-ring appearance. Many of these cells have ruptured, and the 
fat has escaped into the interstices, where it has formed little pools. 
Invasion of the capsule and septa by the tumor cells has occurred. No 
intravascular growth is found, however. Scattered through the stroma 
are also many mast cells, none of which contain oxidase. 

Sections from the areas of ''mucoid degeneration" show an abimdanoe 
of loose wide reticulum containing mucin. In this are present a modern 
ate number of tumor cells of all types with large and small fat droplets. 
With specific stains mucin can be demonstrated in most of the cells 
in the form of a fine or coarse granular deposit in the cytoplasm, clus- 
tered about the fat droplet^.^ Mucin is present in small amount in the 
ceUs of various other parts of the tumor but is most abundant where it 
exists in macroscopic quantities. 

DISCUSSION 

A lipoma is a very common and benign tumor composed of 
mature fat cells. A liposarcoma is much less frequently met 
with and in a large percentage of cases has shown areas of 
myxomatous degeneration. Knox (1) in 1919 collected twenty- 



112 VICTOR C. JACOBSON 

five cases and added two more. The names lipomjrxosarcoma, 
myxoliposarcoma, liposarcoma myxomatodes, and myxosarcoma 
lipomatodes have been used by various writers, depending 
upon which tissue was in excess, the fatty or mucoid, or whether 
the writer regarded the mucous tissue as a product of degen- 
eration of the fatty tissue or vice versa. "Sarcoma" denoted 
invasion, recurrence, or metastasis. Fibrous tissue, cartilage, 
and bone have also been found in these neoplasms, making 
them rather mixed tumors, although with aU elements derived 
from the same germ layer. Hirsch and Wells (2) studied a 
retroperitoneal timior of 69 pounds, parts of which were pu ye 
lipoma, others fibrosarcoma. In none of the cases described) 
have mitotic figures been found in the fat cells, although evi- 
dence of direct division has been noted. The most frequent' 
sites of these timiors are the retroperitoneal tissues, the muscu- 
lar fasciffi of the thigh and leg, the mesientery, the cheek, and 
the breast. 

The tumor forming the basis of this paper has several unique 
features. The bulk of it is composed of cells in practically all 
of which is much fat. Most of these cells are of the embryonal 
tyi>e, being spindle or stellate with delicate fibrillary processes. 
Many transitional forms are present, up to and including the 
large round signet-ring mature fat cell. Many cells are in mitosis 
BSkd these also contain fat droplets. About one-fourth of the 
tumor is bathed in mucin, and the cells in such places contain 
both fat and mucin. 

In offering an explanation for the presence of mucin and fat 
in the same cell, we must go back and consider the embryo- 
logical development of the connective tissues. The connective 
tissues which include fat and mucous tissue are derived from 
the mesenchyme. Todd and Bowman (3), in 1845, were among 
the first to suggest that adipose and areolar tissues are distinct 
and independent. KoUiker, in 1856, described the original fat 
cells as glands, such being found in the axillae and mesentery, 
and around the kidneys. In 1870 Toldt's (4) investigations 
led him to the conclusion that the ^'glands" of Kolliker are the 
source of fat, and he called them ''fat organs." Flenmung 



Fia, 1. pHOTOORAPH OF THE Orioinal Tdmor, Exterior 



Fia. 2. Photograph of the Tumor, Opened by LoNniTUDiNAL Incision 
The bulk of the tumor ia pale pink, opaque and firm. The crosaea (x) murk 
nrctie of myxiimAtous tissue. 



I 
I 



114 VICTOR C. JACOBSON 

(5) in 1876, in a comprehensive study, advanced the opinion 
that the ancestral tissue of fat is a fibrillar connective tissue. 
Hammar (6) also held this view. 

In 1894, Borden (7), in a comparative study of the fat cell, 
stated that (i) in the lower vertebrates the fat cell is developed 
from one form of cell arising from special centers; (ii) in the 
higher vertebrates the fat cell is developed from two forms of 
cells differing greatly in size and shape; and (iii) one of the 
original fat cell forms in the higher vertebrates is homologous 
with the special center forms of the lower vertebrates. This 
cell is gland-like and in no way resembles the connective-tissue 
cell; the other form, while closely resembling the connective- 
tissue cell, is most probably a special cell derived from the 
special center forms by cell division and ^ ^migration metastasis.'* 
Multiplication is by indirect division and the cells never turn 
into connective tissue cells. 

Bell (8), working with material from a calf fetus and from 
steers, described a peculiar open-meshed *'pre-adipose'' tissue 
which forms well defined lobules before true fat cells appear. 
This '^pre-adipose*' tissue consists of loosely arranged cells 
with two or more coarse processes. The branched cells may 
contain fat droplets a long time before they assume the rounded 
form. Bell supports Flemming in the belief that the pre- 
adipose tissue is clearly a fibrillar connective tissue. He also 
described how the cells nearest the blood-vessels filled with 
fat first, the processes becoming absorbed. Altmann's granules 
were observed when the cells were yet branched and before the 
first fat droplets had appeared. 

In Lewis and Stohr's Textbook of Histology (9) it is stated 
that in the four-month embryo the fat cells are quite like the 
surrounding fibroblasts, being fusiform or stellace and con- 
taining vacuoles or droplets. Mallory (10) emphatically says 
that the fat cell is a perfectly definite type of cell formed by 
differentiation from a mesenchymal cell, that it is not a fibro- 
blast, does not arise from one, and, when it undergoes atrophy, 
does not turn into one. 



STUDY OF A LIP0MTX08ARC0MA 116 

The earlier writers, in speaking of the ccfnnective tissue cell, 
probably referred to what is now termed the fibroblast. Conse- 
quently, at the present time the opinion is strongly held by 
' many observers that the fat cell is essentially a modified fibro- 
^ blast, whereas others assert that it cannot be so considered. 
I beUeve that the demonstration in this tiunor of mucin in 
the cells which otherwise appear as fat cells supplies an argu- 
ment in favor of the former idea. 

Mucin is a slimy substance produced by certain epithelial 
(beaker) cells and by the fibroblast under special conditions. 
We are concerned here with its production by the latter cell. 
The umbilical cord furnishes the classical example of mucous 
tissue. The mucin is said to be secreted by the fibroblasts 
without the formation of special granules or vacuoles (Lewis 
and Stohr). Cohnheim restricted the tenn mucin to those sub- 
stances which are excreted by epithelial cells, and reserved 
the name mucoid for closely allied bodies which occur in various 
organs of the body, such as the vitreous humor, cornea, umbili- 
cal cord, and tendons. 

Mucin has certain characteristic chemical and physical 
reactions which permit its identification, so that the term 
mucoid seems unnecessary except for describing gross physical 
qualities of the fluid. It is not produced normidly in the adult, 
but in tumors of mesenchjnnal origin it is found not infre- 
quently, as in myxoma, fibroma, chondroma and osteoma, 
and in the more rapidly growing myxosarcoma. Ewing (11) 
states that mucous tissue ''is genetically related to fat tissue 
into which it is extensively transformed during normal growth. 
• . . . It does not appear that true myxomas ever tend 
to differentiate into fibroma or lipoma.'' He also asserts that 
the spindle and star cells in the mucous matrix may contain 
hydropic or fatty droplets. 

The question naturally arises, is the substance in the tumor 
cells which stains with Mayer's mucin stains really mucin? 
The tissue in which these cells are foimd in greatest numbers 
is bathed in a slimy fluid which gives the accepted tests for 
mucin and likewise stains with mucicarmine and mucihematein. 



tl6 VICTOR C. JACOBSON 

It is, of course, impossible to be certain that the intracellular 
substance is mucin, yet it stains ' 'specifically'' and any other 
interpretation would seem illogical. The cells otherwise have 
ihe appearance of young fat cells such as are found in other 
parts of the tumor. 

It might seem unwise to attempt to settle the origin of a cell 
solely by study of a tmnor derived from that type of cell, particu- 
larly in view of the frequency of metaplasia in pathological 
experience. But metaplasia apparently has a very narrow scope, 
its best illustrations being in epithelial reactions, and applica- 
tion of the principle to the connective tissues has led to great 
differences in interpretation. To explain the secretion of mucin 
by the cells of this tumor as an example of metaplasia does 
not seem as rational as to assimie that the cells which are 
essentially fat cells take on the characteristics of a less differ- 
entiated tissue, a noteworthy deviation from the normal being 
that the mucin is stainable in the cells. 

These observations, it seems to me, indicate that the fat 
cell and the fibroblast (considering -^he mucous connective- 
tissue cell as a modified fibroblast) are very closely related and 
give support to the hjrpothesis that the fat cell is derived from 
the fibroblast. 

REFERENCES 

(1) Knox: Proc. N^w York Path. Soc, 1919, xix, 90* 

(2) HiitscH AND Wblls: Am. J. M. Sc, 1920, clix, 359. 

(3) Todd 4MD Bowman: Physiol. Anat., 1845, i, 90. 

(4) Toldt: Quoted by E. A. Schaafer, Textbook of Microscopic Anatomy, 

New York, 1912, p. 130 (Quain's Anatomy, vol. ii, pt. 1). 

(5) FLbmming: Arch, f.'mikr. Anat., 1S76, xii, 434. 
(OXHammab: Arch. f. mikr. Anat., 1895, xlv, 512. 

(7) Bobdbn: New York M. J., 1894, lix, 225. 

(8) Bbu.: Am. J. Anat., 1909, ix, 401. 

(9) Lbwis ANd StOhb: Textbook of Histology, 2d ed., Philadelphia, 1917, p. 62. 

(10) Mallobt: Principles of Pathologic Histology, Philadelphia, 1914, p. 84. 

(11) Ewing: Neoplastic Diseases, Philadelphia, 1919, pp. 166, 167. 



I 



PLATE > 



/ 



117 



PLATE 1 

Fio. 3. Drawing of microscopical field, high power. Frozen section stained 
with Scharlach R and hematoxylin. The distribution of fat in the cells and 
interstices is shown. In the center are two cells in mitosis and containing fat 
droplets. 

Fig. 4. Drawings of four ceils in a section stained with Scharlach R and 
hematoxylin, showing the transition from spindle cell to the mature signet-ring 
fat cell. 

Fig. 5. Drawings of three cells in a section stained with Mayer's mucicarmine, 
from an area of myxomatous tissue. The finely granular mucin is seen in the 
cytoplasm between the fat droplets. 



118 



STUDY OF A LIPOMYXOSARCOMA 

VICTOR C. JACOBhON 



PLATE I 




I 










K 



rt*;\ 




I 



lY 





119 



PARABIOSIS AND TUMOR GROWTE 

ISIDOR EROSS, M.D. 
From Coluinbia University, Institute of Cancer Research, F. C. Wood, Director 

Received for publication June 15, 1021 

The subject of tumor immunity is of considerable importance, 
and any approach that promises to throw light upon this as yet 
unsolved problem should be thoroughly investigated. The effect 
of parabiosis upon tumor growth has been studied by various 
investigators, but their results and conclusions have been far from 
uniform. In an attempt to clear up this matter, the experi- 
ments here reported were imdertaken. 

Sauerbruch and Heyde (1), in a series of experiments with 
parabiotic animals, noticed that the death of one animal was 
always followed by the death of the other in a few hoirns unless 
the live animal was cut away from its dead partner. This, 
they maintained, is probably due to the absorption of cadaver 
toxins from the dead animal. They also noticed that when 
iodine solution was injected into one animal, the urine of the 
other gave a positive iodine reaction within forty-five minutes. 
That corpuscular elements also could pass from the one to tibe 
other, they proved by injecting a culture of anthrax bacilli into the 
right animal and recovering the characteristic organisms from 
the heart's blood of the left. From their anatomic studies, the 
authors conclude that the operative wound uniting the two ani- 
mals heals exactly as does that in a single animal, except that 
the reaction is more intense, owing probably to ^Hhe foreign 
body" reaction. 

Friedberger and Nassetti (2) studying the antibody formation 
in parabiosis found that agglutinins formed as a result of the 
injection of typhoid bacilli or Vibrio elbensis into one animal 
could be recovered from the non-injected one. They also demon- 
strated that one animal could be immunized passively by the 

121 



122 ISIDOR KBOSS 

injection of these organisms into the other. Morpurgo (3), in 
a series of most interesting experiments, showed convincingly 
that after a bilateral nephrectomy was performed upon one rat, 
the kidneys of its partner in parabiosis were able to compensate 
for their loss and maintain life for both animals — ^in one instance 
for forty days. In another case, he succeeded in keeping two 
parabiotic rats alive for seventeen dajrs after unilateral nephrec- 
tomy in one animal, the two kidneys of the other having been 
previously removed. Jehn (4) and Sauerbruch and Heyde 
obtained similar results in rabbits. 

The findings in all these investigations show how very intimate 
is the biological relationship between the parabiotic animals. 
Lambert (5) studied the influence of parabiosis upon the growth 
of transplantable tumors in rats, and concluded that the 
growth of mouse tumors in rats is definitely promoted by 
parabiosis in that the percentage of successful inoculations is 
increased, the rate of growth accelerated, and the duration of 
active growth extended. Rous (6) in a series of experiments 
with rats concluded that parabiosis has no effect upon the 
growth of tumors. Albrecht and Hecht (7), however, showed 
that parabiosis in mice exerted a distinct retarding influence 
on the growth of tiunors, and explained this phenomenon as 
similar to the immunizing action of subcutaneous injections of 
blood, embryo, or normal tissue. 

In a previous study (8) the author showed that the intraperi- 
toneal injection of blood from immune animals into susceptible 
ones was without retarding influence upon the growth of tumors 
in the latter. This paper is a continuation of the previous 
study and is based on the following working hypothesis: We 
assume that an inmiune animal possesses substances antagonistic 
to tumor growth, or that it lacks those elements necessary to 
tumor for its growth. Having proved that these substances 
do not reside in the circulating blood, we further assume that 
they exist somewhere else in the organism. By performing a 
parabiosis between a refractory and a susceptible animal, 
we bring these two organisms into very intimate biological' 
relationship. We have fiu'ther assmned that it may be possible 



PARABIOSIS AND TUMOR GROWTH 123 

(1) for the ^'antibodies" in the immune animal to cross over into 
the non-immune parabiotic partner and confer upon it the 
property of antagonism to tumor growth; (2) for the substances 
that make for tumor development in the susceptible animal to 
cross over into the immune one and render that animal suscepti- 
ble. To test these possibilities this study was undertaken. 

In one part of these experiments, 44 pairs of animals were 
united in parabiosis. Two distinct breeds of animals were 
employed — ^August rats, which are susceptible to the growth of 
Jensen rat sarcoma (J R S) and Marshall rats which are refrac- 
tory to this tumor. Five days subsequent to the uniting of two 
of these animals, the Marshall rat was inoculated witii 0.003 
gram of Jensen rat sarcoma. Of the 44 pairs, 3 died before 
the tumor inoculation. The parabiotic life period of the others 
varied from four to thirty-five dajrs. Five pairs were lost leaving 
36 pairs from which to draw conclusions. In only sixteen instan- 
ces (44 per cent) could tumor formation be made out and at that 
the majority of these growths were pinhead size. Of the control 
animals, individual rats of the same breed receiving similar inocu- 
lations of Jensen rat sarcoma, 32 out of 43 (74 per cent) showed 
positive results. These tumors were larger, appeared earlier, 
and lasted longer than those in the parabiotic series. 

In chart 1 are shown the tracings of the growth of the Jensen 
rat sarcoma in the Marshall rat. The gradual growth soon 
reaches its peak and in a short period of time the tumor recedes 
and disappears completely. In the parabiotic Marshall rats, 
in practically all cases where the tumor does develop, it fails to 
grow as well as in the controls and recedes much more rapidly. 
In the first two instances on chart 2 it should be noted that the 
rapid growth took place only after the August animal had died. 
The obvious conclusions are (1) that the susceptible animal does 
not transmit any of its characteristics to the immune one; (2) 
that parabiosis reduces the rate, intensity, and duration of 
growth of the Jensen rat sarcoma in the Marshall rat. This is 
probably due to the ill health of the animal in parabiosis. 

In a second series, tests were made to determine whether 
protective substances are transmitted from the immune to the 



124 



I8IDOB KBOS8 



auficeptible animal, in other words, whether passive jmmunity 
can be developed in this manner. The susceptible Marshall 
rats were inoculated with Flexner rat carcinoma (FRC). Nine 



JISlHTO NAtSMa 



IIATIt.10 



M 



It tt 



tf 



# 4 



# # 



# • 



I • 

Z 



J IS IVIt NAtMAU 
MAT » W M 



It 



UnUMTtD 

• # • 



NAT H ilIlK I • It It tl tt tt JVUr 3 



NAT It t* 



IT 






HAT! 



tl 
tl 



t tt 

t • 



tl 



to JUNE4 f 

• • •• 

tCMtATCD 



tt'w 



CbABTS 1 AND 2 



pairs were operated upon; six pairs survived the four day post- 
operative period and were inoculated with the timior. |Two 
survived for a period of over two weeks, which allowed chartings 



PARABIOSIS AND TUMOR GROWTH 



125 



enough to make a fair comparison with the control. Chart 3 
shows that the growths in time of appearance, and rapidity of 
development are only slightly behind those of the controls. 



FIC IMTO NAMNAU 



MIAII«tl9 

MMMUU AM AIIMtT 



«. ^ .. t OCVill.lt 

. . • t 



• 



• • 



• ^ T tcf Jt.tf 

-. . — t •€» 

4 

FIC im NAItNAU COITML 



• • • 



• • 



• • • 

• f f f 




10 CH 



ChABTB 3 AMD 4 



C0NCLT7BI0NS 



From these experiments it is quite logical to conclude that 
parabiosis does not increase the susceptibility of the immune 
animal nor does it give immunity to the susceptible one, and that 
whatever the substances are that make for growth in the suscepti- 



126 I8IDOB KB088 

ble and for non-growth in the immune animal, these properties 
are not carried over by the parabiotic union. The lesser size of 
the tumor and the lesser rapidity of its growth in the susceptible 
partner can be explained by the setback that the operation has 
produced and by the interference with the normal activities of 
the animal, and it is unnecessary to assume as do Albrecht and 
Hecht that this phenomenon is one of partial artificial immunity 
similar to that produced by the injection of homologous tissue, 
since such immunity is known not to affect tumors after growth is 
begun but only to prevent inoculation. While the number of 
animals employed in these experiments is rather small, the results, 
taken in conjunction with tiiose of other observers in this field, 
are sufficiently consistent to make it unnecessary to repeat them 
in a larger series. 

REFERENCES 

(1) SaubbbbuchandHbtdb: Mttnchen.med.Wohii8clir., 1006, Iv, 153. 
^) Fbixdbxbgbb and Nasbxtti : Ztschr. f . Immunit&tsfonch. u. exper. Therap., 
1909, ii, 509. 

(3) MoBPTTBoo: Verhandl. d. deutsch. path. Gesellaoh., 1910, ziv, 259; 1909, 

xiii, 150. 

(4) Jbhn: Ztschr. f. exper. Path. u. Therap., 1909, vi, 16. 

(5) Lambbbt, R. a. : J. Exper. M., 1911, xiii, 257. 

(6) Rous: J. Exper. M., 1909, xi, 810; Proo.Soo. Exper. Biol, and Med., 1909-10, 

vii, 12. 

(7) Albbbcht and Hbcht: Centralbl. f . allg. Path., 1909, xx, 1039. 

(8) Ebobs: J. Cancer Research, 1921, vi, 25. 



THE PROTEIN CONTENT OF THE WHOLE BLOOD AND 

PLASMA IN CANCER 

RUTH C. THEIS 

From the HunHngton Fund for Cancer Reaearehf Memorial Haefrital, and the 
Harriman Beeearch Laboratory, RooeeveU Hoepitalf New York 

Received for publieation August 1, 1921 

In 1910y Dr. William S. Stone and the writer (1) reported a 
study of the chemical composition of the blood in cancer cases, 
describing the results of determinations of the sugar and the 
non-protein nitrogen constituents of the blood. It has been 
a question for years whether the proteins of the blood are in- 
creased or decreasedin cancer. Only recently Robin (2) stated that 
the serum protein is increased, but, like so many other investiga- 
tors, he based his conclusions on the study of only a few bloods. 
Therefore, it seemed advisable to investigate a sufficient number 
of cases and compare the results with those obtained in other 
diseases. 

The nitrogen of the whole blood, serum, or plasma was deter^ 
mined by the Kjeldahl method, and the hemoglobin with Dare's 
hemoglobinometer. Cancer bloods were obtained from the 
wards of the Memorial Hospital, and the others from the Roose- 
velt Hospital. 

Forty-three bloods were examined. Tables 1 and 2 show 
the results obtained on cancer bloods, the only difference being 
that table 1 refers to serum and table 2 to plasma. Table 3 
contains the results on bloods from other pathological conditions. 

The protein content of the whole blood depends to such a 
large extent upon the amount of hemoglobin present that we 
should expect a large variation. Thus, in table 1 whole blood 
protein varies from 12 to 19 per cent, with an average of 16.4 
per cent. 

127 



128 



RUTH C. THEIS 

TABLE 1 
Protein of whole hlood and serum in cancer 







WHOLB 






NT7MBBB 


DIAGNOSIS 


BLOOD 
PBOTBIN* 


SBBUM 
PBOTBIB 


DABB 
HBIfOGLOBOf 






peretnt 


percent 


percent 


1 


Papilloma of bladder 


12.1 


7.7 


40 


2 


Carcinoma of bladder 


17.6 


6.6 


80 


3 


Carcinoma of bladder 


16.2 


6.0 


90 


4 


Carcinoma of prostate 


19.0 


7.3 




6 


Carcinoma of prostate 


19.1 


6.7 


80 


6 


Carcinoma of prostate 


12.1 


6.9 


65 


7 


Carcinoma of prostate 


16.2 


8-2 




8 


Carcinoma of parotid gland 


17.8 


6.9 




9 


Carcinoma of uterus 


14.1 


7.1 


51 


10 


Carcinoma of rectum 


18.1 


6.4 


90 


11 


Epithelioma of jaw 


16.2 


6.0 


85 


12 


Epithelioma of tonsil 


16.8 


6.6 


75 


13 


Neurosarcoma of arm 


18.2 


7.7 


76 



^Nitrogen X 6.25 » protein. . 

TABLES 

Protein of whole hlood and plasma in cancer 



HVMBBB 


BIAaXOSIS 


WHOLB BLOOD 
PBOTBIlf* 


PLASMA 


DABB 
BBMOOLOBIB 






percent 


percent 


percent 


14 


. Carcinoma of bladder 


16.9 


7.1 


76 


15 


Carcinoma of bladder 


13.2 


5.3 


60 


16 


Carcinoma of bladder 


16.1 


5.6 




17 


Carcinoma of bladder 


18.8 


7.1 




18 


Carcinoma of uterus 


15.5 


5.6 


60 


19 


Carcinoma of breast 


14.7 


6.1 


76 


20 


Carcinoma of breast 


15.7 


5.9 




21 


Carcinoma of rectum 


14.4 


8.1 




22 


Carcinoma of rectum 


15.0 


5.9 




23 


Carcinoma of liver 


15.8 


6.1 




24 


Carcinoma of laiynx 


14.7 


6.7 


1 


26 


Carciiioma of stomach 


8.7 


5.1 




26 


Teratoma of testis 


18.2 


6.7 




27 


Keloid of chest 


19.3 


7.4 





^Nitrogen X 6.25 » protein. 

In table 2 the protein varies from 13 to 19 per cent with an 
average of 16.3 per cent, if we except case no. 25, in which the 
blood was very anemic with a protein value of only 8.7 per cent. 



PROTEIN CONTENT OF BLOOD AND PLASMA IN CANCER 129 



In table 3 the protein varies from 12 to 20 per cent with an 
average of 17.6 per cent. Hemoglobin readings were, unfortu- 
nately, not done in these cases but they would undoubtedly 
have averaged higher than in the cancer cases and would account 
for the slightly higher protein value of the whole blood. 

If there is an increase or decrease in the amount of protein 
in cancer blood, it would show in the serum or plasma. Here 
we can see that the variation is very slight. Table 1 varies from 
5.9 to 8.1 per cent with an average of 6.8 per cent; the patient 

TABLES 

Protein of whole hlood and jjiUuma in other diaeaeee 



KUIIBSB 


DXAGNOSia 


WBOLB BLOOD 
PBOTBZN* 








ptreetU 


per cent 


1 


Hematoma of kidney 


19.8 


7.2 


2 


Tuberculous glands 


16.4 


5.4 


3 


Colon bacillus infection 


18.2 


5.8 


4 


Paralysis agitans 


19.6 


7.9 


5 


Pleurisy 


15.3 


6.0 


6 


Ear infection 


19.6 


8.0 


7 


Retroversion of uterus 


16.0 


6.9 


8 


Urticaria 


17.9 


8.1 


9 


Possible adhesions 


19.6 


7.6 


10 


Neurasthenia 


17.8 


7.2 


11 


Neurasthenia 


17.3 


7.4 


12 


Pyelitis 


18.4 


8.0 


13 


Pyelitis 


12.3 


7.9 


14 


Pyelitis 


20.0 


5.9 


15 


Pyelitis 


18.4 


7.2 


16 


Pyelitis 


16.3 


8.0 



* Nitrogen X 6.25 « protein. 

whose blood had the highest value died the day after the blood 
was taken. The plasma protein values of table 2 varied from 
5 to 8.1 per cent, with an average of 6.3 per cent. The plasma 
values of other pathological conditions varied from 5.3 to 8 
per cent, with an average of 7.0 per cent. 

Loebner (3) reported results on 44 cases of cancer. She 
employed the refractometric method, using serum obtained from 
finger blood. Her figures varied from 4.2 per cent to 9 per cent 
with an average of 7.3 per cent, which corresponds very closely 

TBB JOT7BNAL OF CAlfOBB BBBBA.BCB, VOL. TI, NO. 2 



130 RUTH C. THEIS 

with the average of this series. She considered 8.2 per cent the 
average for normals since that is the average which Heudorfer 
obtained in a study of five cases. This would indicate that 
the protein of the serum is somewhat reduced in cancer. 
If, however, we compare the average in cancer cases with the 
average in cases of other pathological conditions, we can see 
that there really is no difference in the amount of protein present. 

CONCLUSION 

Proteins of the blood plasma are neither decreased nor in- 
creased in cancer cases as compared with other hospital patients. 

REFERENCES 

(1) Thbib and Stone: Jour. Cancer Res., 1919, iy, 349. 

(2) Robin: Bull, de TAcad. de m^., 1920, Ixxxiv, 51. 

(3) Lobbnbb: Deutsoh. Arch. f. klin. Med., 1918, ozxvi, 61. 



PROBLEMS IN CANCER RESEARCH^ 

MONTROSE T. BURROWS 

From the Department 0/ Surgery, Washington UnioerBity School 0/ Medicine and 
Research Laboratories of the Barnard Free Skin and Cancer Hospital, 

Saint Louis, Missouri 

Received for pabliofttion June 6, 1921 

It is a pleasure to announce that a Department for Cancer 
Research has again been established in St. Louis by the Barnard 
Free Skin and Cancer Hospital. Hie material available is the 
laboratories and what can be developed from the more general 
biological side within .them, and the clinical material. 

The plan which is being used for the study of cancer divides 
the work into several distinct groups. One is a general study of 
the properties of the various cells of the body and the peculiarities 
of the mechanisms of growth, division, differentiation, and func- 
tion of these cells. A second is a close study of the clinical 
course of the disease and the results of methods of treatment 
and diagnosis. A third is an attempted analysis of isolated 
facts now fully demonstrated, and their immediate relation to 
clinical and pathological data at hand. 

For the immediate time, one problem which seems to be 
most pressing is the improvement of methods of diagnosis, 
especially the diagnosis of internal cancer. Early diagnosis of 
cancer of the breast has prolonged the life of many patients for 
at least ten or twenty years. The same is true for other super- 
ficial growths. Barring a few exceptional cases, methods for the 
diagnosis of internal cancer have not reached the same degree of 
perfection, as, for instance, cancers of the intestinal canal. If 
by chance a cancer of the intestine causes early obstruction, a 
diagnosis may be made and the treatment may be hopeful. How 
few cancers of this region give, however, this early symptom I 

^ PreMnted before the American Aseooiation for Cancer Researohp Cleveland, 
Ohio, March 24, 1921. 

131 



132 MONTROSE T. BUKROWS 

Another of the iminediate problems is the betterment of 
methods of treatment now available. The third and most im- 
portant of all is the etiology of this disease. It is most impor- 
tant, because through its elucidation and even in its study all 
other problems must become simplified. Through it alone, 
barring some peculiar accident, can the hopes for ultimate 
success in treatment come. The greatest effort in cancer must 
be exerted, therefore, in this direction. 

What is known about cancer today, aside from clinical con- 
siderations, is little more than was known many years ago. This 
is appalling but true, and the question arises: Is it not due to 
the fact that the general notions of the cell, its structure and 
properties have not materially advanced during this time? The 
problem of the etiology of cancer when resolved into terms of the 
cell may be quoted thus : Is the cancer cell a cell different from 
normal cells, or is it a normal cell suffering a continuous external 
growth stimulus? Since the application of the theories of cellu- 
lar growth now in vogue have failed to answer this question, 
there remain open but two channels of attack; the hit and miss 
method, and a further probing into the nature of the mechanisms 
of cellular growth, division, differentiation, and function. 

Cancer, no matter how it is taken, represents a break in the 
nonnal growth-regulating conditions of the body. Without a 
knowledge of the normal growth-regulating mechanism, it is to 
be expected that the etiology of cancer will remain obscure, unless 
it is proved that the disease is the result of some definite foreign 
agent or stfanulus — ^parasite, or otherwise — ^which may be isolated. 
It is known that the body is composed of cells. The minute 
metabolism of these cells, the mechanism of their formation, even 
the simplest constituents of their life are, however, based only on 
theories. The leucocytes are compared to the amoeba. One 
talks glibly of cellular metabolism, when he probably means 
body metabolism. Is this so-called cellular metabolism the 
work of the cell or is it the work of the organ or the body as a 
whole? In beginning a few years ago the study of the cells in tissue 
culture, it was of interest to note that the leucocjrtes, both the va- 
rious mononuclear as well as the polymorphonuclear types, show 



PBOBLEMS IN CANCEB RE8EABCH 133 

no ability to grow and divide like other cells. Their movements 
are passive reactions to external conditions. A connective tissue 
cell or a heart muscle cell suffering from the effects of certain 
toxic poisons can be made to resemble the small and large mono- 
nuclear cells of the inflammatory exudate. These latter cells 
have lost, also, all ability for further growth imder the condi- 
tions in which they had previously grown or under any other 
conditions it has been possible to impose upon them. In actively 
growing culture many of these fixed tissue cells, imaffected by 
foreign poisons may also show such changes after they have 
been preyed upon by actively growing cells. 

The polymorphonuclear leucocjrte is a cell differentiated in a 
special locality. Lymphocytes tmdoubtedly are foimd in the 
same locality as well as in lymph-nodes. Whether the large 
nimiber of mononuclear cells of the chronic inflammatory area or 
about cancers have this origin is still a very much open question. 
In fact, the work of Maximow has not been confirmed. (See 
Stockard (1).) That these cells contain enzymes and bacterio- 
lytic substances which are liberated when tiiey break down is, 
again, well established. What is not known are the conditions 
under which these cells live in the areas about a cancer and 
whether they do liberate these substances imder ordinary condi- 
tions there. It seems evident, therefore, that they loight as 
well be considered the result of active growth or the presence 
of toxic substances, as to be thought of as antagonists to these 
conditions. In chronic inflammations, it is generally where the 
bacteria are most active that they are most prevalent. In the 
tissue culture it is the actively growing tissue cells that show the 
true antagonistic action to growing bacteria. The lymphocytic 
cells and the rounded-off connective tissue cells intermingle in 
the cultures with the bacteria. The connective tissue cells, 
the heart muscle cells, and other actively growing cells cease 
to grow when they come within the zone of bacterial action. 
The bacteria stand them off as they also stand off the bacteria (2). 

This absence of an absolute knowledge of many of the most 
important functions of body cells must in itself make it very 
unsafe, therefore, to draw any far reaching functional conclusions 



134 M0NTR08S T. BT7RB0WS 

from morphological data. It makes it absolutely necessary to 
draw one 's conclusions from the clinical facts alone. Hiis applies 
not alone to treatment, but also to other experimental results. 
The laboratory in Saint Louis is well situated in this regard in 
that it is a definite part of the clinic, and also allows the broader 
type of biological research which in the light of present knowledge 
must form an intimate part of the cancer laboratory. 

PBOBLBMB I7NDBB CONBIDEBATION 

1. Studies on oxidation in actively growing and differentiated 

cells f espedaUy in reference to their growth 

In a previous communication (3) before this society evidence 
was given by the author to show that connective tissue cells 
and simple mesenchymatous and embryonic muscle cells show 
no evidence of an internal organization suitable for growth. 
They act more like homogeneous fluid systems, leaving aside 
the nucleus and the astral center. Growth in them can be 
explained entirely as a surface tension phenomenon, the energy 
of the process being effected by the formation of a substance 
within the cell, which is insoluble in circulating body fluid, but 
soluble in dead cells, fibrin, and products of di^tegrating cells. 
A cell forming this substance and brought in contact with fibrin 
suffers a decrease in surface tension along the line of contact. 
The further details of the organization peculiar to growth is 
determined by these contacts. The order of the contacts of 
a growing and dividing heart muscle cell is different from a 
rhsrthmically contracting one. 

For a long time it has been known that actively growing 
tissue cells, such as those taken from yoxmg embryos and 
sarcomata commence their growth in the culture within one or 
two hours or much earlier than adult tissues. This is seen in 
a study of tissues of embryo chicks. The cells from a frag- 
ment of the heart of a two, three, or four days old chick embryo 
will commence to grow almost at once when placed in the culture. 
About fragments of a ten day old embiyo heart such activity 
is not observed imtil after six or twelve hours, and about frag- 



PBOBI^MB IN CANCEK BESBABCH 135 

ments of a fifteen day old chick embryo heart after fifteen or 
twenty hours; while about young adult heart fragments this 
latent period is often as long as twenty-four or forty hours. 

In a previous publication (4), the writer reported a study of the 
relation of oxygen to these changes in the cell. A method was 
developed which allowed accurate regulation of the amount of 
this gas in the chamber about the cells. The tissues especially 
studied were fragments of the heart and body wall of fourteen, 
fifteen, and sixteen days old chick embryos. It was found that 
the growth was as active in an atmosphere of 9 per cent oxygen 
as in pure oxygen^ and ceased in atmospheres which contained 
slightly less than 6 per cent of this gas. It became of interest 
this year to see whether the more actively growing tissue might 
not show variations in these figures. The question that arose 
was whether the less actively growing tissues did not have to 
form or in some other way liberate the substance or substances 
which actively change surface tension or produce energy. The 
short latent period in the more actively growing cells suggested 
a supercharging of this substance or substances. 

In 1903, Fletcher (5) showed that a muscle of the adult frog 
will give maximum contractions every five minutes for a period 
of two hoiu^s in an atmosphere of pure nitrogen. In the work on 
organization of isolated rhythmical, contracting heart muscle 
cells in the tissue cultiu-e, the author has found evidence to 
show that the chemical changes of muscular contraction are 
similar in certain stages to those of growth. If muscle will 
contract for a time without oxygen the question that arises is: 
Will not cells fully differentiated for growUi show the same 
reaction (6)? 

The tissue so far tested has been cells from the hearts of chick 
embryos of various ages, ranging from four to fifteen days. 
Fletcher 's experiments were repeated with contracting hearts and 
heart muscle fragments of these embryos isolated in the tissue 
culture. These cells will contract for from twenty to twenty- 
two hours in an atmosphere of nitrogen. In the same oxy- 
gen-free atmosphere, the cells from fragments of four and five 
days old chick embryos will grow from four to six hours. This 



136 MONTROSE T. BtmROWS 

growth commences, however, always after a considerably longer 
latent period than that in an atmosphere containing oxygen. It 
is active for this short time, after which the cells invariably 
disintegrate rapidly. The period of growth in these hanging- 
drop ciiltm*es is very much longer in air; the early disintegration 
is rarely or never seen; and when it occurs, it is always later. To 
obtain a similar growth about the fragments of the heart of 
ten daya old chick embryos 1.8 per cent oxygen was required and 
about fragments of fifteen days old chick embryos 4.5 per cent. 

These experiments have not only been applicable to the prob- 
lem in question, but have given a method of differentiating func- 
tionally the actively growing young embryonic cells from those 
showing greater degrees of differentiation. In the case of these 
cells they have also shown the necessity of Gxyg^n for the main- 
tenance of structure. Whether this method may be used for 
differentiating any actively growing tissue is a problem yet to 
be considered. 

STUDIES OF PIANT CANCEB8 

Another method which has appealed to us in attacking the 
cancer problem has been the study of plant cancers. These 
represent an active proliferative growth, the etiology of which is 
known. Chambers (7) has xmdertaken this problem. He was 
interested during last summer in investigating the metabolic 
activity of the organism. Bacterium tumefaciens, in culture, 
and in the course of this study noted that these organisms pro- 
duce alkali in the culture even in the presence of sugar, which 
they also cause to disappear rapidly. Smith, by a different 
method, had noted that ammonia is formed in the cultures of 
these organisms. Harvey (8) more recently, also, noted that the 
tissue of these tumors is more alkaline than normal plant tissue. 
The work of Chambers would indicate, therefore, that the in- 
crease in the OH-ion in the tissue of these tumors is due to the 
etiological agent, which produces alkali in the cultures at least. 
Chambers is now attempting to see if under the condition imposed 
on them by the plant, they also form it. 



PB0BLEM8 m CANCEB RE8BARCH 137 

These facts attracted our attention to the work of Menten 
(9) I by which an alkalosis was demonstrated in the blood of 
cancer patients; and this question is now under investigation. 
The various methods of testing were critically reviewed and the 
indicator method was selected. The tests are made on a dialy- 
sate of venous blood, which is taken under oil and kept imexposed 
to air throughout the whole procedure. The results so far ob- 
tained have been interesting. In the early part of the disease, 
these patients show no evidence of change in the reaction of 
their blood when it is taken from a region far removed from the 
cancer. A definite alkalosis is observed only when the cancer 
has grown to considerable extent and involves neighboring 
glands and tissues. The blood of a patient with a beginning 
carcinoma in the lip is normal. A definite H-ion change in the 
blood is not seen until the glands of the neck are involved, and 
then increases gradually as the disease progresses. The same is 
true for carcinoma of the intestine. A small annular carcinoma 
of the intestine^with no or but little glandidar involvement may 
show by the indicator method no change in the alkalinity of the 
blood. Sharp changes are always seen, however, when liver 
nodules are present. The COs relations in the blood of these 
patients is now being studied. It is possible that by this means 
much concerning the significance of this alkalosis may be deter- 
mined. The anemia may also have an effect. 

The importance of these experiments is still to be ascertained. 
That the alkali is probably produced in the malignant tumor is 
indicated, however, in the study of one patient suffering from 
a sarcoma of the forearm. Blood taken directly from a vein 
leading from the tumor showed a pH of 7.5, while in the other 
arm the blood pH was 7.3, or normal. This case illus- 
trates the difficulty of using such a method for a clinical test. 
The body compensates more readily for acid than alkali, but at 
the same time there is a rapid elimination of alkali by the kidney. 
In the case of acid, the compensating mechanism gives evidence. 
To what extent this can be used for determining the presence of 
alkali is yet to be detennined. In the case cited, there was 
evidently alkali being given into the blood by the tmnor, but it 



138 MONTROSE T. BUBROWS 

wapS not detectable in the other arm. There was no glandular 
involvement in this case. 

What the further studies of this reaction will open up and the 
further studies of the bacteria of the plant tumors may reveal 
is a question which the future alone can answer. It seems, how- 
ever, that this method may yield something important. 

REFERENCES 

(1) Stocxabd, R.: An experimental analysis of the origin and relationship of 

blood corpuscles and the lining cells of vessels. Proc. Nat. Acad, of 
Sc, 1915, i, 666. 

(2) BxTBBOWB, M. T. : Unpublished notes. 

(3) BuBBOWS, M. T. : The tissue culture in cancer. Proc. Second Pan American 

Sc. Cong., 1916-1916, Sec. viii. Pt. 2, pp. 494^196. 

(4) ByBBOWB, M. T. : The oxygen pressure necessaiy for tissue activity. Am. 

J. Physiol., 1917, xlui, 13-21. 
(6) Flbtchbb, W. M.: The influence of oxygen upon the survival respiration 
of muscle. J. Physiol., 1917, xxviii, 854-369. 

(6) BuBBOWB, M. T. : The reserve energy of actively growing tissues. Proc. Soc 

Exper. Biol, and Med., 1921, xviii, 133. 

(7) Cbambbbs, W. H. : Paper before the Am. Physiol. Soc, Chicago, December, 

1920. 

(8) Habvxt, R. B. : Relation of catalase, oxydase and H-ion concentration to 

the formation of overgrowths. Am. J. Botany, 1920, vii, 211-221. 

(9) MxNTBN, Maud L.: The alkalinity of the blood in malignancy and other 

pathological conditions; together with observations on the relation of 
the alkalinity of the blood to barometric pressure. J. Cancer Res., 
1917, ii, 179-211. 



THE INFLUENCE OF HEREDITY IN DETERMINING 

TUMOR METASTASES* 

BTUDIES IN THE INCroENCE AND INHERITABILITY OP SPON- 

TANEOUS TUMORS IN MICE 

Sixteenth Report 

MAUD SLTE 

Fmm the Cancer LohonUory of the Otho 8. A. Sprague MemarM InHUuU and tk§ 

Univeraily of Chicago 

Received for publication August 1, 1921 

Tumor metastases in this stock are somewhat rare. In 29,000 
necropsies^ furnishing something over 4000 primary spontaneous 
tumors, mostly malignant, only about 19 per cent of the tumors 
have metastasized. It has been the general opinion of patholo- 
gists that the formation of metastases is a distinguishing property 
of malignant tumors. The inGltration of tumor cells into 
tissues adjacent to the neoplasm, and the entrance of these cells 
into the lymph and the blood streams are the usual mechanical 
bases for secondary growths. But aU these conditions exist in 
innumerable instances, both in human and in animal tumors, 
without the occurrence of secondary growths. For example, 
females 7821 and 9260 of this stock (see chart 11, strain 508, 
p. 158) each had a large and highly malignant infiltrative mam- 
mary gland carcinoma of relatively long standing. In both 
instances there were multiple tumor emboli in the lungs, but no 
secondary growths arose from these emboli, — ^i.e., the mechanical 
basis was there in each case, the occurrence of tumor did not 
follow. 

^ Presented before the American ABflociation for Cancer Research, Cleveland, 
Ohio, March 24, 1021. 

139 



140 MAUD SLYB 

Ewing (1), in his chapter on metastasis, speaks of ^'a peculiar 
susceptibility of some tissues to develop secondary txmiors," 
but later states that ''the mechanism of circulation will doubt- 
less explain most of these peculiarities, for as yet there is no 
evidence that any one parenchjmotatous organ is more adapted 
than others to the growth of embolic tmnor cells." There 
would seem, then, to be some factor, other than the mechanical 
one, which determines the incidence and location of metastasis, 
even though it be minimized, as Ewing has done in the para- 
graph cited. 

Again, Ewing states that in highly vascular tissues like the 
lip, stomach, and testicle, very small carcinoma may yield 
distant metastases.* But these very organs are among those 
whose tiunors rarely metastasize in this stock. For example, 
among the lip carcinomas in this stock none has metastasized; 
among the testicle tiunors (2), one metastasized locally, while 
five out of six stomach carcinomas (3), (4), showed metastases 
in regional lymph-nodes only, but no vascular metastases in the 
liver or elsewhere. And this in spite of the close resemblance 
of these tumors to human tumors of similar type in similar 
organs. 

In this stock, mammary gland carcinoma and sarcoma rarely 
metastasize into regional l}rmph-nodes, in marked contrast to 
human breast tumors of the same tjrpe. Indeed, in this stock, 
Tnfl.TnTn5i.ry gland timiors can be seen growing up to the lymph- 
node but not invading it, even by extension. On the other 
hand, pulmonary metastases from manunary gland carcinoma and 
sarcoma are common in this stock, again in contradistinction to 
the reported behavior of hmnan breast tumors of the same types. 

These divergencies (which are tjrpical of the many divergen- 
cies) in the metastasis behavior of tiunors of similar types and 
in similar organs, require some explanation other than merely 
the mechanical tendency of certain types of tumors to metas- 
tasize in certain locations, and that there is some other factor is 
generally recognized by pathologists, although little progress 
has as yet be^ made in demonstrating what this factor is. 



HEREDFTT IN DETERMINING TUMOB METASTASES 141 

The matter of tumor metastasis in this laboratory has been 
under observation for twelve years, and the data here given 
are based on over 29,000 necropsies, among which every instance 
of metastasis is known. The charts used to demonstrate the 
points made in thj/s report are perfectly typical. The non-tumor 
fraternities (where there are any) hdve been omitted, in order to 
save space and to show escactly where these tumors occurred. 

It must be borne in mind in reading these charts, that, as I 
have previously stated, tumor metastases in this stock are 
rather rare, being only about 19 per cent of the primary malig- 
nant tumors. A given chart then, which shows many primary 
tumors, must not be expected to show lai^e numbers of secondary 
tumors. For example, a chart showing 8 primary tumors, 
would, if this general average of 19 per cent were carried out, 
show only 1.5 cases of metastasis, etc. The biological evidence 
from this laboratory, which explains this low percentage of 
metastasis from malignant tumors, will be given further on in 
this report. 

From this study three points stand out with striking clearness: 

POINT I 

In any given strain the metastatic tumors (where there are any) 
tend to occur in exactly the same organs in which the primary tumors 
of that strain occur. 

Note chart 1, showing part of strain 215 and some of its deriva- 
tives. The parent female, no. 3, had a sarcoma-carcinoma of 
the mammary gland, a malignant adenoma of the liver, and a 
metastatic sarcoma of the kidney. The parent male, no. 360, 
came from a strain carrying limg and mediastinal tumors, and 
he himself was proved heterozygous to tumors in these locations 
(i.e., he carried them potentially and introduced them in every 
strain into which he was crossed, although he did not himself 
show these tumors). 

Note here how these particular types and locations of tmnors 
have '^ segregated out'' and certain of them have been trans- 
mitted to the branches of the strain here shown, both as primary 



142 



MAT7D SLTB 









K 



ui 

> 

i 



u 

o 

bl 

t 
O 
10 



I 

M 
Z 
< 










HBREDITY IN DETERMINING TX7M0R METASTASES 143 

and as secondary neoplasms. In the first hybrid generation, 
male 1192 had a primary carcinoma of the lung and a secondaiy 
carcinoma of the lung. Female 5 had a primary carcinoma of 
the mammary gland, with multiple secondaries in the limg. 
Note how the different types and locations of timiors have 
'' segregated out/' so that in branch A l\mg tumors alone have 
been transmitted, aud a 100 per cent Ixmg tumor strain is being 
extracted. In branch B liver tumors have predominantly 
been transmitted, from female 630 with a secondary tumor in 
the liver, so that in one part of the branch a 50 per cent liver 
tumor strain is being extracted. 
In these two branches of the strain, then, female 3 introduced: 

a. Primary carcinoma of the mammary gland, in females 5, 
2239, and 630. 

b. Primary adenoma of the liver, ia females 1681, 8871, 
11237, and 12207, and in male 2327. (Note the three consecu- 
tive generations of adenoma of the liver.) 

c. Secondary carcinoma of the liver, in female 630. 

d. One case of sarcoma, in female 5543. 
Parent male 360 introduced: 

a. Primary carcinoma and adenoma of the lung, in males 
1192, and 900, and m females 326, 816, 823, 3191, 2548, and 6688. 

6. Secondary carcinoma of the lung, in male 1192 and in 
females 5 and 630. 

c. Primary tumor of the mediastinum, in female 5543. 

d. Secondary tumor of the mediastinum, in female 326. 

Note how female 5543, with a lymphosarcoma of the medias- 
tinum, derived the type of her tumor from her maternal ancestor 
(female 3) and the location of her tumor from the paternal 
ancestor, male 360 (who carried these tumors potentially, and 
transmitted them in every strain into which he was crossed,' 
although he himself did not develop them), just as in the human 
species, a child might inherit curly red hair from a mother with 
curly black hair, and a father with straight red hair. 

Note how female 5 derived her primary tumor from the 
maternal ancestor, and her secondary timior from the paternal 
ancestor, and how female 630 derived her primary tumor from 



144 MAUD SLTE 

the parent female, the secondary in the lungs from the parent 
male, and the secondary tumor of the liver from the parent 
female — a marked demonstration of the ''segregating out" of 
these things, which is ^ heredity test. Note how, after the 
second hybrid generation, no further mammary gland tumors 
occurred in either of these two branches. 

Note that there occur no tiunors, either primary or secondary, 
which were not introduced by the parents. Note that the 
secondary tumors occur in the same organs as the primary 
tumors. 

Chart 2 shows parts of strains 48 and 292. These are two 
strains derived from the same original mating as was strain 
215, that is, female 3 (with a sarcoma-carcinoma of the mam- 
mary gland, a malignant adenoma of the liver, and a metastatic 
sarcoma of the kidney) , and male 360 who introduced carci- 
noma of the lung and the mediastinimi. Here again the particu- 
lar types and locations of tumors introduced by the parents 
have "segregated out" and been transmitted to the resulting 
strams, both as prunaiy and as secondary neoplasms. 

In the first hybrid generation female 5 derived from the 
mother a carcinoma of the mammary gland, and from the 
father a secondary carcinoma of the l\mg. In the second hybrid 
generation, female 37 derived from the parent male a carcinoma 
of the lung, with a secondary carcinoma in the mediastinmn. 

In the third and fourth hybrid generations, note the out- 
cropping of sarcoma^ both primary and secondary, in females 
1454, 26, 348 and 396. Liver tiunors derived from female 3, 
also occur in these strains, as shown in females 5743, and 26, and 
in male 11836. There is also a " pre-cancerous " liver in female 
399. Here also lung tmnors, derived from male 360, occur 
primary in females 37 and 399 and in male 11836; and secondary 
m females 5, 5743, and 396. 

Note the occurrence of secondary carcinoma of the mediasti- 
num in female 37, of secondary sarcoma of the mediastinum in 
female 396; of secondary carcinoma of the lung in female 5, and 
of secondary sarcoma of the lung in females 5743 and 396. Note 
the three cases of secondary sarcoma in the spleen in females 



HBBEDITT IN DETBBMININO TUMOR METASTASBS 



145 



6 



•I 
M 

Q 

I 

Z 
< 

a: 
I- 



I 
i 



I 

•5 ® 






fc; 



J 



I 



o 

3 « 



o 



u I ^ 



Oh 



1 



i 

40 



O Cm 



O 



1. 

40 



2*^ 

it 



I 



1 

Of 



9^ 



i 






in 


I 



^1 



o*> 



o 



s 



1 

J 



.<o 




J 



3l 



s; 



o 





<o 



l\ 



Jl 



h 



o 



I 



m 
t 

ii 

o 



C9 

H 



I 






JOUBITAL or CAHCIB 



TOb. TI, WO. 2 



146 MAXTB SLYE 

5743, 348, and 396. Rarely in this stock, except in strains 
derived from female 3, have there been any tumors of the efpleen 
either primary or secondary. 

Here again, after the first hybrid generation, mammary gland 
tumors disappear— another indication of how these things 
'^ segregate out." The mammary gland tumors preponderated 
in other strains derived from the same original mating. The 
dominating timiors, both primary and secondary, in these 
strains are in the liver, kidney, spleen, mesentery, Itmgs, and 
mediastimun. 

In these strains, then, female 3 introduced: 

a. Primary carcinoma of the mammary gland, in female 5. 

h. Primary adenoma of the liver, in female 399 and in male 
11836. 

c. Primary sarcoma of the liver, in females 5743 and 26. 

d. Primary sarcoma of the kidneys, in females 26, 348, and 396. 
6. Secondary sarcoma of the kidnejrs, in female 1454. 

/. Secondary sarcoma of the spleen, in females 5743, 348, and 
396. 

g. Primary sarcoma of the mesentery, in female 1454. 

h. Secondary sarcoma of the mesentery, in females 1454, 
5743, and 26. 

Parent male 360 introduced : 

a. Primary limg txmiors, malignant and not yet malignant, 
in females 37 and 399, and in male 11836. 

&. Secondary lung tumors, in females 5743 and 396. 

c. Secondary carcinoma of the mediastimun, in female 37. 

d. Secondary sarcoma of the mediastimun, in female 396. 
Note here again, how, though they are much fewer, the second- 
ary tmnors arise in the same organs as do the primary. 

Chart 3 shows a part of strain 65, branch 2. The parent 
female, 5738, had three carcinomas of the mammary gland 
and metastatic ttunors in the lungs. She was mated with her 
brother, male 782, who died of uncertain causes, but who was 
proved heterozygous to tmnors of the mammary gland (he 
carried them potentially and transmitted them, but did not 
himself develop them). Note here that with one exception 



HBREDITT IN DBTERBONINO TTJMOB METASTASES 



147 



s 

< 
o 



s 

I 

s 

z 

I 




m 
Z 

X a 

9 ♦ 



5 



o 



o o 
Of 



9 
2 
3 
-I 

i 

O M 
S OB 

i :: 



I- 



•r 5 i 



J 

o 



O 



o 





< I 

Jl 



o» 



fO 



I 
5 



O 



1 


« 


s 


« 




<» 


< 


lO 




Of 


1 




J 




1 




e 
< 






Of 



I 



148 MAX7D 8LYE 

(female 9888 who had a lymphosarcoma of the thymus) the 
only tmnors which occurred were carcinoma of the mammary 
gland and limg tumors, primary and secondary. 

Note the five consecutive generations of carcinoma of the 
mammary gland in branch B, all but females 5753 and 9903 
showing lung metastases. Note the heavy percentage of lung 
tumors primary and secondary in both branches of the strain. 
Note the negative evidence of the '' segregating out" of the 
different types and locations of tumors, both primary and 
secondary, in the complete absence from this strain of all tumors 
of the liver, kidney, mesentery, spleen, pancreas, uterus, etc. 

Chart 4 shows a part of strain 338, branch V, A, with partial 
ancestry. The parent female 3920 had an adenoma of the liver. 
Note the predominance of primary liver adenoma in this strain, 
7 cases (over 43 per cent) and the occurrence of two cases of 
secondary liver timxor in the same strain, that is, female 8865 
with an osteoid sarcontia of the mammary gland, secondary in the 
liver, and her grandson, male 16370 with an osteosarcoma of the 
leg, metastasizing in the liver, and also a mesothelioma of the 
testicle. (Testicle timiors also occur in this strain.) 

Chart 5 shows a continuation of the same strain, strain 338, 
branch V, A (the data being too extensive for a single chart). 
Chart 5 begins with generation 3 of the family, the mating being 
between female 8619 and male 8751. Generations 7, 8, and 9 
are added to the family in this chart. Note here, also, the 
further occiurence of liver tumors, there being one additional 
primary, in female 30469, who showed a primary liver adenoma 
in addition to seven carcinomas of the mammary gland, a pri- 
mary lung adenoma, a typical uterine fibroid, and a diffuse 
endothelial hyperplasia of the glands and spleen. (Needless to 
say, the mouse was riddled with tumor, little but neoplasm 
remained.) There were abo in this part of the family, two 
additional cases of secondary liver tumor, one carcinoma and one 
sarcoma, that is, female 22263 with three carcinomas of the 
mammary gland and a sarcoma in the peritoneum near the 
stomach, with sarcoma metastases in the liver, pancreas, and 
uterine wall, and female 30501 of generation 9, with two carci- 



HBBBDITY IN DBTESSMININO TUUOB METASTASES 



149 



IE 

i 



4 



«• 

u 






?5 

J s 
<< '^ 

•o' 



r 



•lO b 



o»- 







o 



o»- 



z 



J* 



*r 






J.50 



MAUD SLTE 



nomas of the mammary gland and the liver absolutely riddled 
with carcinoma secondaries. 

It is interesting to note here also the two occurrences of uterine 
neoplasm, primary in female 30469, and secondary in female 
22263, as this is one of the strains canying most of the uterine 
timiors m this stock. 



CONTIMUATION OF STRAIN 338— BRZ 



o * 



2Caiu P) G«. 

— r 



AoC««U>MA UtVIR 



aCAac ^.Gi. 



9| 



A»*MOWWI LlVBH. 

lOftfO 



Caac t^ Cl 

A»»M4MM n.&IU 

lOOhh 



1 



LI 1^0 



2 CmcM. Gt, 
Saac M Ok. 
laabi 



* ITC8I ^ 



I 



Cahc Lwm« 



<S* 



P. LlUKKMIA 



3 CaWL M Gc. - Pi*lTONtA«. Sam.c. 
MtTA* LivgK. - PanCIU.A« 



?L 



ICaivc m Ck 



SAaCOAAA HftAP 

AMD NtCK 



I 



JCaRC.M. Gl -' Livk* Ao«lttOA^A 

LtMO ADBMo/w\«t — U-raaiNe FiaMe.o 
Dirrwti EMDOrH. Hvrc«.pi.A«iA ar 

J0#fc9 GkAND* AHO SVkBlN 



Chabt 5 



•2. CAA.C. M.Ct 
i.<va«. MKAKt.r 



l»r A\«TAV 



JO^0A 



09 



The pancreatic secondary sarcoma also is extremely interest- 
ing, as this is the strain which carries the only other pancreatic 
neoplasms which have occurred in this stock. 

The complete ancestry of this strain is shown in charts 6 and 
7, chart 6 being the ancestry of parent female 3920, a;nd chart 7 
the ancestry of parent male 3024. The prominence of liver 
tumors, primary and secondary, in the resulting strain is due to 
the fact that it was bred in from both sides of the family; the 



I 



HESEDITT IN DETEBMINING TUMOR MBTASTA8E8 



151 



ANCESTRY OF Q 39Z0 -" STRAIN 35S 



<l»Mui«r 



I 



Cm«c. Smk. H.Ci. 

Mala. Aocnoma Livkb ^ 

3 Samc^cta* Kidmcv O 



90 



L 



CaucMGi. 
1 



J 



6UP. RM40PTMAL,MITi» 
lfc7 



I 



OmcMGl 

791 O 

— ^^^^^ X ■ 



iNTf STINAk InFCCTIOM 
1481 



Advnoma Livcr 
3920 



^nucrMT 



Chabt 6 

strain thus received ''a double dose/' Note also the marked 
outcropping of sarcoma in strain 338, derived from original 
parent female 3. Sarcoma also was bred in from both sides of 
the ancestry. 

ANCCSTirr or 6 302a -stkainms 



Canc SmicKOi.. 
3 5am. ^Arnk%. Km 



Cmmom Nbm4MitU 
30 



Amnoma Uvkr 
863 



Unccktain 
842 



J 



L 



441 6 



1011 



J 



IChMN. N/lpHOITIft 
lHrk.AM.LuMC NOoukB 
U 1^084 



C. 



ntuMT 



Chabt 7 



Chart 8 shows a part of strain 465, which was derived from 
the same original ancestry as strain 338 (ancestry shown in 
charts 6 and 7). Note here also, the same appearance of liver 
tumors, two primary and one secondary. .This gives us, then. 



152 



MAUD SLTB 



in these two small strains (338, branch V, A and 465) derived 
from identical original ancestry, 10 primary and 5 secondary 
liver tumors. Including the ancestry, the number is 13 primary 
and 5 secondary liver timiors. This is a very high occurrence 
of liver tumors, both primary and secondary, especially when one 
considers the fact that outside of this stock there is just one 
mouse liver tumor on record, a primary tumor of the liver, re- 

PART OF STRAIN 4€5 



"MUtSMT 



IAocroMA LivfR 
^ 3920 



Cmron. NtPM«iri* 



6 <3024> 



i 



5333 6 

— r — 



Unccrtaim 



2512 



SmICOM* LlVCR 

5305 



Chiio(«.Ncph. 
5215 



ACCIOCNT 

5ftS3 



J 



Umccm 
6369 



SMtcHtNoLca 
Rm rmo9K Sakc - A^cta« 
At,mMT i«sn.Ac<Ne 
9979 . *-•*** 



L 



Uncertain 
995T 6 



J 



Cmbon. Nb'mriti* 






9647 6 



ASMCtft LUMO 

8S52 



Lv^PHOio luMOR Stomach 

LvW»HO-&AIICOMA LiVBR. 

Itasa AMoCttMMaw &tiA Duct 



Chabt 8 



ported by the Imperial Cancer Research Fund of England. This 
liver tumor occurrence (in the Slye stock) within a single strain, 
makes a piece of evidence which is conclusive. 

These five strains, represented in charts 1 to 8 inclusive (which, 
it must be remembered, are perfectly typical) furnish both 
posititfe and negaiitfe evidence of the '^s^pregating out'' of differ- 
ent types and locations of tumors, and their transmission as 
such, whether as primary or as secondary neoplasms. They 



HEREDITY IN DBTBBBnNINO TUMOR METASTASES 153 

furnish, therefore, both positive and negative evidence that 
in any given strain ike metastatic tumors {where there are any) tend 
to occur in exacOy the same organs in wMch the primary tumors of 
(hat strain occur. Note, therefore, that: 

1. Every secondary tumor of the liver in this stock (with the 
possible exception of two osteosarcomas, which metastasized into 
nearly every organ in the body, in two mice whose strains have 
not yet been completely analyzed) has occurred in strains carry* 
ing primary liver tumors. 

2. Every secondary tumor of the kidney in this stock has 
occurred pi strains carrying primary kidney tumors. 

3. Every secondary lung tumor in this stock has occurred in 
strains carrying primary lung tumors (with the possible exception 
of the one lung metastasis in strain 164, branch IV, line C). 

4. Every secondary tumor of the mediastinimi in this stock 
has occurred in strains yielding primary tumors of themedias- 
tiniun. 

5. Every secondary tmnor of the pancreas in this stock has 
occurred in the strain which yields primary pancreatic tmnors. 

6. Every secondary tumor of the spleen in this stock, has 
occurred in strains carrying primary spleen tmnors. 

7. Every secondary uterine tumor in this stock has occurred 
in strains carrying primary uterine tumors. 

When this stock, or any other, has been completely analyzed, 
there is little doubt that it will show that every secondary ttmxor 
of any organ whatever has arisen in a member of a strain which 
has shown, or will show, some cases of primary tumors in that 
organ (however few they may be). There is little doubt that 
strain 164, branch IV, line C, of this stock will in time yield a 
primary Ixmg tmnor, although it has not done so yet. There is 
a very slight inheritance of lung tiunor neoplasm tendency in 
this line of the strain, and it is therefore to be expected that 
lung timiors of any kind will occur but seldom. 

The complete analysis of any strain is a very long and difficult 
process, and the apparent testimony of the frequent occurrence 
of secondaries in man in tissues where primary tumors rarely 
occur is of no help here, as no human strain has ever been even 
partially analyzed. 



154 MAITB 8LTE 

Female 3 is listed in these chai-ts as having a sarcomarcarci- 
noma of the mammary gland. This tumor was unquestionable 
carcinoma in some parts and apparently sarcoma in other parts. 
There has been considerable difference of opinion concerning 
these mixed tumors. Slye, HolmeS; and Wells (5) Loeb (6), 
Lewin (7), Bashford (8); Woglom (9), and some others, diagnose 
theim as such, while Le Count (10) and Ewing (11), etc., consider 
them to be entirely carcinoma, with pressure distortion of the 
cells simulating sarcoma in some areas. 

The biological evidence of the work in this laboratory un- 
questionably supports the opinion that there are these mixed 
tumors. Female 3 with such a txmior, has unquestionably 
transmitted carcinoma in every strain into which she has been 
hybridized; and she has equally certainly transmitted sarcoma 
in some branch of every strain into which she has been crossed 
(see charts 1, 2, 4, 5 [female 3 shown in ancestry charts 6 and 7], 
8 (same parentage charts 6 and 7), Chart 9). 

I have preferred the term sarcoma-carcinoma to represent 
this type of tumor, because according to the biological evidence, 
it is not a sarcomatous carcinoma, nor a carcinomatous sar- 
coma, but rather a sarcoma plus a carcinoma, each type of tumor 
being transmitted separately, as such. Biological evidence, as 
manifested in heredity behavior, is too fundamental to be ignored. 

POINT II 

In certain strains there is a tendency for primary tumors to 
metastasize in certain organs; whereas, in other strains, tumors of 
the same type, primary in the same organ, even when they are of 
older and of larger growth, fail to metastasize into those organs. 

Chart 9 shows part of strain 73, which was derived from the 
same female 3, mated this time with male 30. Male 30 came 
from a strain carrying tumors of the lung, mediastinum, and 
diaphragm, and was proved heterozygous to tumors of these 
organs, having been tested in various crosses. 

The son of this mating, Jap. W. F. male, died before autopsies 
were made, and consequently the cause of his death is not known. 



HEBEDITT IN DETBRMIKINO TUMOR MBTASTASEB 



155 












V 



m 
Z 
I 

ft 




s 
i 



3 




?.- 



I 



40 



0(> 



1 

11 



o 



II 



O 



1 



i 






Of 



o 
z 






58 
£4 






i 






«o 



» 






Of 



11 

fi 



f 



Of 



to 



II 



Of 






t 9 



13 






♦O 






< « 



40 



i 

5 



o- 



V? 



156 MAUD SLTB 

He was crossed with a Lathrop silver-faun female, who also 
died before autopsies were made. Their offspring, however, are 
shown in two branches, A and B. 

Note how in branch B, every mammary gland carcinoma 
metastasized into the lungs; even the retroperitoneal and pelvic 
sarcomas (male 3951) metastasized in the lungs. Note also 
that with ihe exception of male 1775 this branch shows 100 per 
cent of lung txunors, primary and secondary. Note the two 
generations of metastasis into the diaphragm, female 1722 and 
male 3098. 

On the other hand, in branch A, carcinomas of the same type 
and of older and larger growth, failed to metastasize into the 
lungs (females 4328, 5723, and 7544). Note that in female 
5723, although the mechanical basis for Ixmg secondaries was 
present in midtiple timior emboli throughout the lungs, no 
lung tumor developed. Note that the sarcoma of the mesen- 
tery in male 4630 metastasized into the liver, kidney, and retro- 
peritoneal glands, but not into the lungs. 

Note the three consecutive generations of leukemia in this 
strain, in males 899, 1785, and 2360; and the further fact that, 
although the lungs are one of the principal organs for leukemic 
infiltration, in males 1785 and 2360 there was no infiltration in 
the limgs, while in male 899 the lung infiltration was very slight. 
On the other hand, the marked leukemic infiltration in this 
family was in the same organs in which the tumors of the parent 
female occurred, that is, the liver and kidney. The metastatic 
tiunors., also, in this family were in these same organs. 

In branch B, however, the 100 per cent Ixmg tumor strain, 
female 6568 who had pseudoleukemia along with an adenoma 
of the lung, showed marked infiltration of pseudoleukemia 
throughout the lungs. 

We have here, in these two branches of the same strain, a 
striking illustration of point II : In branch B all timiors metas- 
tasize into the lungs, and pseudoleukemia picks out the limgs 
predominantly; while in branch A, the tiunors of the same type 
in the same organ and of older and larger growth fail, in every 
case, to metastasize into the lungs. Even the leukemic infiltra- 
tion fails to take hold of the lungs in this branch. 



HBREDITY IN DETBBMIKINO TUMOR METASTASES 



157 



Chart 10 shows part of strain 405. In this branch of the 
family; after the parent generation, although there are five 
straight generations of mammary gland carcinoma, none of these 
tumors metastasized into the lungs. Note also the complete 
absence of primary lung tumors. This branch of the family 



G. 



PAREMT 



G.. 



G., 



o.. 



G. 





PART OF 5TRAIN 405 


5 


3 Carc. /^.Gl. 
A\ETA5. Lungs 
5738 


6 


Uncertain 
78a 




f 




9 


Car.c.M.Gl. 

CHRON.f^ASTtTW. 

5753 

1 K 

1 


6 


PuL. Infection 

2715 


9 


Garc M.Gl. 

55ai 

1 


6 


Chron.Nepm 
Pbritonitis 

4449 




1 






9 


Carc. M.Gl. 
Tapcworia 

1 


6 


Wounds 

t571 


9 


Carc. M.Gl. 
9053 

1 


6 


Wounds 
10342 


9 


Carc.M.Gl. 

Chabt 10 





G. 



RARKNT 



G., 



Gv, 



C. 



did not inherit lung timiors, and none occurred, either primary or 
secondary. 

Chart 11 shows one branch of strain 508. Here again through 
five consecutive generations of manunary gland carcinoma, no 



158 



MAUD SLTB 



lung metastasis occurred. Even where the presence of multiple 
tumor emboli in the limgs gives certain evidence that cells from 
the primary tumor were carried to the limgs (in females 7821 and 
9260) no lung ttunor occurs. The mechanical basis was there; 
the tumor growth did not follow. This strain also shows no 
primary lung tumors. 

PART OF STRAIN-508 



<*-PAReNT 



G.. 





2 CARC.ri.Gu 




Occlusion of stomach 




^ 


59fc5 


6 


t9GS 


^PARENT 














CARC.rA.GL. 

Tumor Emboli Lunos 


6 

1 


Chron. Nephritis; 




^ 


7621 


■f 




5212 


ftt 












Carc.MGl. 




\fj0UND& 




9 


7237 


t 


701H- 


«^» 












$ 


CarcHGl. 
Tumor Emb 
92tO 


OL) Lur 


4Gb 


6 


Accident 

7i*60 

1 


S 






- 






2 Carc. MGl. 






^ 


9989 






««.- 






Chabt 1 


I 




T 



Chart 12 gives a summary of strain 164, branch IV, families 
A, B, and C. It is shown in summary only in order to get the 
data within a single chart. These branches of the family showed 
48 primary tumors, all but one of them highly malignant, with 
only one case of lung metastasis, in female 10574, a member of 
line C. Note the marked absence of lung tumors primary and 
secondary, although lung tumors are second in frequency of all 
mouse tumors. 



HBBBDITT IN DBTBRMININ6 TUMOB METASTASES 



159 



The strain also showed 13 cases of pseudoleukemia and one of 
leukemia. Here again, it is most interesting to note that the 
six cases of pseudoleukemia which occurred in line B showed not 
the slightest infiltration in the lungs. The malignant thymus 



STRAIN Ifei#.-BII.1S:.A-B-C 


PRODUCED THE. TOLLOWINO CASES OF TUMOR 


WITHOUT A\ETA!>TASIS 


WITH METASTASIS 


C/\RCINO/v\A M.Gl. J«h 


1 IN LUfslG 


5arco/v\a /^.Gi.. 2 


O 


OVAMAN AOBNOAAA , ^ 


O 


A\ALlO. TnyMus Lympmoaaa 2, 

DID NOT iNFliTTRATE LUNO-S 


o 


LrMPHO- Sarcoma thymus t 

919 NOnr INriUTKATK UUNG-ft 


o 


5q.CtJ-i. Cakc. Back 1 


o 


3p. Ceui. 5arc. 0AeK ^ 


o 


5<\.Cci.i. Carc Face 1 


o 


5p OfSUX. 5ARC.rAC£ 1 


o 


Retropfritomfai. Sarcoaaa 1 


o 


Total 48 


1 


PSruDO Leukemia It 




Note the absence or PRIA^ARY lung TU/^AOR AND 

THE SINGLE Case of lung AAETASTASIS 



Ghabt 12 



tumors in this line of the family also failed to infiltrate the limgs, 
although they went extensively into the chest wall and medias- 
tinal glands. The seven cases of pseudoleukemia which fell in 
line C of this strain, showed a very slight infiltration of the lung. 
This is'tiie same line which showed one case of limg metastasis. 



160 MAX7D SLTB 

We have then, in this strain, lines A and B, a marked iUus- 
tration of the f aihire of the lungs to yield at all to tumor, primary 
or secondary. In other words, the lungs of this family do not 
show neoplastic proliferation, nor do they accept neoplastic 
secondaries or leukemic growth in spite of the certain and marked 
presence of the ordinary provocation for such proliferation. In 
line C of the strain, a sli^t tendency of the limg to proliferation 
is present. 

This strain was derived by the hybridization of an absolutely 
nontiimorous strain of house mice, in my hands many years, 
strain 358, with albino strain 146, which carried many tumors, 
and furnished some 100 per cent tmnor families. This strain 
gives another marked illustration of the segregating out of these 
things; the tendency to neoplastic lung proliferation, either 
primary or secondary, failed to be transmitted at all to lines 
A and B of branch IV of the strain, and only slightly to line C. 
Note the striking contrast between this strain with no Itmg 
tumors, and strain 522 (chart 15) showing 100 per cent Ixmg 
tumors (p. 167). 

These four strains represented in charts 9 to 12 inclusive 
(which are perfectly typical) furnish both positive and negative 
evidence of the segregating out of different types and locations of 
tumors, and then- transmission as such, whether as primary or as 
secondary neoplasms. 

They furnish therefore evidence that in certain strains there is 
a tendency for tumors to metastasize into certain organs ^ whereas in 
other strains, tumms of the sam^e type in the same organ, even when 
they are of older and of larger growth, fail to metastasize into those 
organs. Moreover in these strains, leukemia and pseudoleuke- 
mia tend to infiltrate the same organs which show pnmary and 
secondary neoplasms, and fail to take hold of the same organs 
which do not yield to primary or secondary tumors. 

In this connection, it is interesting to analyze the results of 
various experiments to test the effect upon metastasis formation 
of making exploratory incisions into tumors, of massaging 
tumors, etc. Let me indicate what would be a complete biolog- 
ical control in such experiments. 



HBRBDITY IN DETEBIONINO TUHOB MSTA8TA6E8 161 



TWO ILLXJBTRATION8 OF COBfPLETE BIOLOGICAL GONTBOU 



Strain 164 branch C of the Slye stock, during its twelve years 
of eidstence has consistently jrielded 0.01 per cent of limg tumors 
primary and secondary, leukemic and pseudoleukemic infiltration 
of the lung, and infiltration into the lungs from malignant thymus 
tumors. 

If now we use timiorous mice from strain 164 branch C, and 
excise for diagnosis pieces of their tumors or massage their tmnors 
or otherwise manipulate them, and if we find at autopsy that the 
group of mice whose tiunors have been manipulated show little 
or no higher percentage than 0.01 per cent of lung tmnor metas- 
tasis, and that the group which is not manipulated runs true to 
this 0.01 per cent lung tipnor metastasis, we shall have a perfect 
control, and shall be justified in saying, without qualification, 
that excising a piece of timior for diagnosis or massaging the 
timior, or otherwise manipidating it, had no efif ect upon the pro- 
duction of lung tumor metastasis. 

II 

Strain 65 branch II B has during its twelve years of existence 
produced 35 per cent of lung tumor metastasis. 

If now we repeat these same experiments with this strain, and 
find at autopsy that the group whose tmnors have had a piece 
excised for diagnosis or which have been in any other way manip- 
ulated, show little or no increase in lung tmnor metastasis over 
35 per cent, and if the immanipulated group runs true to this 
percentage of lung tiunor metastasis, we shall have a perfect 
biological control of the experiments, and shall be justified in 
saying without qualification that the excision of a piece of tmnor 
for diagnosis or any other form of manipulation had no effect 
upon the production of Itmg tiunor metastasis. 



162 MAUD SLTB 

THREE ILLUSTRATIONB OF THESE EXPERIMENTS WITH INADBQTJATE 

BIOLOGICAL CONTROL 



Strain 164 branches A and B have during their existence 
for twelve years, without exception of any sort, yielded per 
cent of lung tumors primary or secondary even where tumor 
emboU were present in the lungs, per cent of leukemic or pseudo- 
leukemic infiltration into the lungs, and per cent of infiltration 
into the lungs by extension from thymus, diaphragmatic, or other 
adjacent malignant tmnors. 

If now we repeat these same experiments with these branches 
of the strain, we shall find at autopsy per cent of lung tumor 
metastasis in every group of mice, whether or not we excise a 
piece of tmnor, or massage the tumor, or otherwise manipulate it. 
That is, a strain whose lungs have been proved not to yield to 
tumor emboU is not a strain fitted to show the effect of these or 
any other types of manipulation. 

II 

Strains 139, 522, 73, branch B, etc., have dining their many years 
of existence without exception furnished 100 per cent of lung 
tumors, primary and secondary, infiltration into the lungs by 
leukemia and pseudoleukemia, and extension infiltration into the 
lungs from malignant thjonus and other adjacent tumors. 

If now we repeat these same experiments with these or any 
other 100 per cent lung tumor strains, we shall find at autopsy 
an exceedingly high percentage of lung tumor metastasis, prob- 
ably 100 per cent in every group whether excised or not. That 
is, a strain which for twelve years has shown 100 per cent of lung 
tumor is not a strain fitted to show the effect upon lung metastasis 
of this or any other tjrpe of manipulation. 

Ill 

Groups of mice even when purchased from one dealer are mixed 
lots, secured from many divergent sources, and wholly unanalyzed 



HEREDITY IN DETBBMINING TUMOR METASTASES 163 

as to tumor potentiality. They may include some individuals 
from per cent lung txunor strains, some from 100 per cent, some 
from 35 per cent, etc. If we take such groups of mice and repeat 
these same experiments, we shall have no biological control, and 
we shall not be justified in saying whether the lung tumor meta- 
stases found at autopsy were or were not caused by the excision 
of a piece of tumor for diagnosis, or by massage, or whatever 
form of artificial manipulation may have been used. 

It is greatly to be deplored that the vast majority of experi- 
mental studies in animal pathology and bacteriology have not 
been and are not as yet bemg conducted with animals whose 
hereditary potentialities have been studied. It is being taken 
for granted that pathological conditions found in the animal 
after experimental procedure, are the result of such experi- 
mental procedure. No allowance is being made for heredity. 

Until stocks of animals to be used in such experiments, have 
been thoroughly tested out as to their inherited potentialities, 
such experiments are lacking in adequate control, as the factor, 
of heredity is not being considered, although it is tremendously 
potent. Heredity alone woxild be sufficient to give the exact 
results attained in countless experiments which have been con- 
ducted without any efifort to study or control the heredity 
factor. Moreover, most of these studies in metatasis production 
by the artificial manipulation of the primary tumor have been 
conducted with grafted tumors. As has frequently been stated 
in the series of cancer studies from this laboratory, the biological 
difference between spontaneous and grafted tumors is so funda- 
mental and so complete, that the behavior of grafted txmiors has 
practically no bearing upon the behavior of spontaneous tumors. 

POINT ni 

IndividuaU with secondary tumors in any given organ seem to be 
as potent as individiuils with primary tumors in the same organ to 
transmit by heredity primary tumors in that organ. 

Note chart 13, showing part of strain 304 with ancestry. Here 
two successive generations of females (529 and 467) with second- 



164 



MAUD 8LYE 



j^ 



z 

s 

i 



o 



6 



(Q 



i 

► 

u 

s 



3 «< 



o 



If 



oT 




V 

< 

I 

t 

IS 

0^ 



1 






* 



I 

1 

\ 






«o 



o 



I s 



*o 






u r ;_ 




H 1 

a 1 


o 


ii 




• 


lO 


£ 


M 


J 


s 




O 


^ 




t 




i 


i. 


u 


^ 




o 


j 


1 




o 



«o 



K> 



CO 

H 




c* 






O 



i 



O 



rf* 



<Sr 



HEBBDITY IN DETEBMINING TUMOR METASTASES 



165 



ary hmg tumors, have originated a strain which, in line C, is 
producing a 44.4 per cent primary lung tumor stram. The 
combination of both lines B and C shows 33.3 per cent of primary 
limg tumors. Note the 100 per cent pseudoleukemia family 
being extracted in branch B, in all of which the lungs were infil- 
trated extensively. 

Agam, in chart 14, strain 338 branch VI, a 100 per cent lung 
tumor strain is being extracted from the original mating of female 

ffAKT or <TKAIN 35S -ftnag 



A^li«^rr 



L 



5«t»7 Ml 



IAdimoma Lwm« 



S 1 7*^3* 



'nikRSNT 



c. 



9i 



2 Cahc m. Ck. 
8419 






O. 



91 



CaaC M.QI.. OM Ta» •» 
16 9 r A\CTA» LUNOS 



S^. Cbl^ Caac. Jaw 



«]' 



07l»9 



Cahc Luno 
5 97« eA«9J 



Caiw./^Ol. 
Mult. Lwmo Mctas. 



CAJtC.^.Ci.. 3Caac.M.Gi.. 

LunoS RiOOkio wn-M |m«ta«. Luw«> | 

12050. MiTAft rktl9<*l 3 1 



9L 



AoaNOAIA 

L.WM« 



XCAf^C.M 6a.. 

MUUT LWM« A<t«TAft. 



2 Caac. M.Ok. 

Mult LwN«M>m% 

15190 Utsiiimb ft9H,Of*^i 



Chart 14 



5417, with a secondary lung tumor, with male 7736^ with a 
primary lung tumor; that is, the metastatic lung tumor is behav- 
ing like a primary lung timior in helping to produce a 100 per 
cent lung tumor strain, primary and secondary. 

Refer again to chart 1 (page 142), showing part of strain 215, 
and note how female 630 (parent of branch B) with a metas- 
tatic carcinoma of the liver, produced three straight generations 
of primary liver adenoma (not yet malignant) females 8871, 11237, 



166 MAX7D SLTE 

and 12207, as well as a primary malignant adenoma of the liver 
in male 2327. Note also how her secondary lung tumor was 
followed by a primary lung tmnor in female 823 (Gs). 

Refer again to chart 2 (page 145), showing strains 48 and 292, 
and note how the secondary sarcoma of the kidney in parent 
female 3, resulted in primary sarcomas of the kidneys in females 
26, 348, and 396. Refer again also to chart 3 (page 147), straiQ 
65, and note how secondary lung tumor in the parent female 5738, 
introduced primary Ixmg tiunors in the strain, in females 1253, 
8749, 11512, and 13386, and in male 11832. Note that the thy- 
mus sarcoma in female 9888 was markedly infiltrating the limgs, 
while in strain 164 Br. IV, lines A and B, which did not produce 
limg tumors of any sort, either primary or secondary and in 
which leukemia and pseudoleukemia failed to infiltrate the lungs, 
the thymus tumors also failed to iofiltrate the lungs. 

Note also how in chart 15, straia 522, female 5636, Vith a 
secondary limg tumor (Gj), mated with male 8102 with a primary 
lung timior, produced a 100 per cent primary lung tumor strain 
(center of the chart) . The pseudoleukemia in this 100 per cent 
lung tumor strain, viz., female 14033, showed marked infiltration 
in the lungs. 

From these charts, viz., 13 to 15 inclusive, and also from charts 
1, 2, and 3 (which are perfectly typical) it is evident that individudU 
with secondary tumors in any given organ seem to be as potent as 
individuals with primary tumors in that organ, to transmit by 
heredity, primary tumors in that organ. 

DISCUSSION 

We have here, in the charts presented in this report, biological 
evidence of the nature of cancer heredity. It is well to bear in 
mind that until the facts of pathology have been based upon the 
more fundamental biological facts, it will never be possible to get 
a complete explanation of the pathological behavior of tissues. 

Let me repeat at this point, what I already have stated fre- 
quently in previous publications: the materials used for this 
study are individuals analyzed as to their hereditary poten- 



HEREDnr IN DBTERMININO TUMOB METASTASES 



167 





• 


' 


cT 


(^ 


nl 


6 






to 
d 


d 


J 










d 

1 • 


h 

X : 

if 
JI 

• B It 

CD J ? 


1 

J s 

1 ! 


4 

f N 




_i 

w 

s 
< 






1 

< 

i 




< 

9 


*o 

Is 

• 

« ? 1 


I*— ^ 




o 
-J 

3 


♦o 

2 




J- 


1 < 


o 

Z 

-J 


Of 

o 

! 

s } 

< - 










♦O 

! 

3 * — 

-J 

Of 

1 

8 ^ 
< S> 


o 

< 


Of 




If 




04-I 




1 f* 




ri 


Of 






f 
^ 


6 

t 

w CO 

< o 


i-^ 


♦0 
.1 ■ 


i"^ 


*o 

r 

;? s 
c3 J 


1 




c J 1 

-1 1 to 

U I 

u< -_ 


I 

8 •> 


io 




Ol- 


J3 


6 i 


? ?£ 




o *0"1 

3 


< t 


t 


w 

01 

lA 

Z 

I 

It. 

o 

1- 

i 


Us 




6ii 




< D 


5 


o*- 


Of 




o 




Of 








Of 
i 

Ji 


•O 

1 

1 9- 

i 1 


o^ 




Of 


O^ 
























i 

CM 2 















Ol- 



I 



168 MATJD SLTE 

tialities by every possible test. The tumors ore all spontaneous, 
arising without interference of any sort except that of selective 
breeding. By the process of selective breeding alone, usmg 
analyzed individuals as material, it is possible to extract strains 
in which the lungs do not react with either primary or secondary 
neoplastic proliferation or leukemic infiltration; or strains in 
which the liver does not so react, or the kidneys, or the mamntiary 
gland tissue, etc. 

Etiological meaning. The fact that both primary and second- 
ary txunors of a given organ or organs tend to occur in the same 
strains, and fftil to occur at all in other tumor strains, indicates 
beyond question that heredity is a strong factor in determining not 
only where the primary tumors shaU occur, but also where the 
secondary tumms shaU occur. 

The consequent fact, that it is possible to extract strains 
where only certain organs shall f lumish both the primary and the 
secondary neoplasms, and shall yield to leukemia and pseudo- 
leukemia; and to extract other* strains in which these organs 
never furnish either primary or secondary neoplasms nor yield 
to leukemic or pseudoleukemic invasion, and that secondary 
tumors are just as potent as primary timiors in the transmission 
of primary timiors in any given organ, shows beyond a doubt 
that the thing which is transmitted in the heredity of cancer is the 
tendency of a given organ or organs to }deld to cancer. This 
inherited tendency of an organ or organs to yield to cancer is 
manifested, whether the lesion is primary in that organ, or 
whether cells from the primary growth lodge in that organ and 
form a secondary lesion. 

The negative evidence on this point is just as conclusive, 
that is, that in strains from which tiunors of a certain organ or 
organs have been eliminated by heredity, the cells from the 
primary neoplasm fail to take hold, even when they lodge in 
such organ, and do not form a secondary growth; and even when 
a timior completely surrounds an organ, it fails to penetrate it 
by extension. 

Now what do these things mean? They mean that the ten- 
dency to carcinoma segregates out and is transmitted as such; 



HEREDITY IN DETERMINING TUMOR METASTASES 169 

that the tendency to sarcoma or adenoma, etc., segregates out 
and is transmitted as such; that a strong tendency to the location 
of one or more of these types of tumor in a specific organ, such 
as the liver, the lung, the kidneys, the mammary glands, etc., 
is transmitted, owing to the segregating out of a peculiar type of 
tissue in these organs, which will respond in a neoplastic manner 
to lesions of any kind which furnish a chronic irritation of not too 
destructive a type. That is, the tissn/es of these organs are of the 
same nature as the ancestral organs from which they are derived. 
There is a specificity of tissue type from liver to liver, or from kidney 
to kidney, etc,, which wiU make those organs react in a given way to 
a given type of irritation. 

Now with each of these things being transmitted as such, it is 
possible for a parent female with a sarcoma-carcinoma of the 
mammary gland, a malignant adenoma of the Uver, and a second- 
ary sarcoma of the kidney (like female 3) to transmit to the 
strains derived from her, a tendency to: 

1. Carcinoma of the mammary gland. 

2. Sarcoma of the mammary gland. 

3. Adenoma of the mammary gland. 

4. Carcinoma of the liver. 

5. Sarcoma of the liver. 

6. Adenoma of the liver. 

7. Carcinoma of the kidney. 

8. Sarcoma of the kidney. 

9. Adenoma of the kidney, or any combination of these nine. 
If she is mated with a male, either himself having a limg and a 

mediastinal tumor, or being heterozygous to these locations of 
tumor (as in the case of male 360) , the resulting strains will show : 

1. Carcinoma of the mammary gland. 

2. Sarcoma of the mammary gland. 

3. Adenoma of the mammitry gland. 

4. Oarcinoma of the Uver. 

5. Sarcoma of the liver. 

6. Adenoma of the liver. 

7. Carcinoma of the kidney. 

8. Sarcoma of the kidney. 



170 MAX7D SLYB 

9. Adenoma of the kidney. 

10. Carcinoma of the limg. 

11. Sarcoma of the limg. 

12. Adenoma of the lung. 

13. Carcinoma of the mediastinum. 

14. Sarcoma of the mediastinum. 

15. Adenoma of the mediastinimi, or any combhiation of these 
tumors. 

The neoplastic growth may be primary in any one or more of 
these organs, accbrding to where there is occasion for the primary 
lesion to occxu:, and there may be secondary growths in any one or 
more of these organs, according to where the secondary lesions 
occur. 

Now note that this is exactly what we get from the matings of 
female 3 (with a sarcoma-carcinoma of the mammary gland, a 
malignant adenoma of the liver, and sarcoma metastasis in the 
kidney) with male 360, proved heterozygous to lung and medias- 
tinal tumors. Note especially charts 1 and 2 where the strains 
were derived from the same original mating of female 3 with male 
360; and charts 4, 5, 6, 7, 8, and 9, showing ancestry and strains 
derived from the mating of female 3 with male 30. 

It should not be a matter for surprise that there is an inheri- 
tance of a specific type of Uver tissue, or kidney tissue, etc., which 
will tend to respond in a given way to a given irritation or a given 
lesion. Similar facts of inheritance, such as a tendency to grow 
tall, with the proper degree of nourishment, etc., are easily 
accepted. That also is the inheritance of tissues which tend to 
respond in a given way to a given stimulus. There is here also a 
certain specificity, aSj for example, the tendency to longness in 
the tnmk and shortness of the limbs; or of ^ortness of the 
trunk and longness in the lunbs, so that a given figure tends to 
predonunate in a given family, just as given types and locations 
of neoplasms tend to predominate in a given family. Again, it is 
an accepted expectation that there should be a given nose shape, 
or shape of lips, or contour of head, inherited within a family; 
i.e., a specificity of nose tissue, etc., which will insure its growing 
to a given shape, etc. 



HEREDIT7 IN DETERBCINING TUMOR METASTASES 171 

This specificity of organ iisstie, which ahaU insure its reacting in 
a given neoplastic way, to a given chronic irritationy is exactly what 
we find to be the nature of cancer heredity; and it is obvious that it 
will make no difference in the nature of the reaction of such 
specific organ tissue, whether the lesion is primary or secondary. 

Heredity, therefore, of a specific tjrpe of organ tissue, is here 
shown to be the fundamental influence in determining the inci- 
dence and location of metastatic neoplasms, as well as those of 
primary neoplasms. 

In regard to the relative infrequency of secondary tumors in 
this stock, the biological evidence is as follows: Many of these 
strains have been made by hybridizing a tumorous individual 
with a proved non-tumerous individual. This, of course, is 
done to test for the Mendehan behavior of cancer in heredity. 
Now, with the tendency to carcinoma, sarcoma, etc., segregating 
out, and a specific type of organ tissue segregating out and being 
transmitted as such, it is evident that any given mouse may 
inherit a tendency to only one type of tiunor, and in only one 
organ. Other organs, therefore, in such a mouse, refuse to 
yield to cancer, even where emboli are present in such organs. 
Many such cases have been pointed out as shown in the charts 
in this report. 

Individuals into whose ancestry (as is the case in much human 
heredity) an indiscriminate amount of cancer had been bred 
would show a correspondingly higher percentage of metastasis. 

8X7MMABT 

These studies in the metastasis behavior of spontaneous 
tumors demonstrate the following facts: 

1. In any given strain, the metastatic tumors (where there 
are any) tend to occur in exactly the same organs in which the 
primary tumors of that strain occur. 

2. In certain strains, there is a tendency for tumors to metasta- 
size in certain organs; whereas in other strains, tmnors of the 
same tjrpe in the same organ, even where they are of older and of 
larger growth, fail to metastasize into those organs. 



172 MAX7D SLTE 

3. Leukemia and pseudoleiikemia, occurring in tumor strains, 
pick out predominantly the same organs for infiltration which 
show the primary and secondary tumors of that strain. 

4. Ttunors do not even invade by extension the organs from 
which primary and secondary neoplasms have been eliminted 
by heredity. 

5. Individuals with secondary tumors in any given organ, 
seem to be as potent as individuals with primary tumors in the 
same organ, to transmit by heredity, primary iumors in that 
organ. 

Therefore : 

6. Heredity is a strong factor in determining' not only where 
the primary tmnors of a strain shall occur, but also where the 
secondary ttmiors shall occur. 

7. Heredity is a strong factor in determining what organs of a 
strain shall yield to the invasion of leukemia and pseudoleukemia. 

8. The thing which is transmitted in the heredity of oan^cer is 
the tendency of an organ or organs to yield to cancer. This 
tendency is manifested whether the lesion is primary in that 
organ, or whether cells from the primary growth lodge in that 
organ and form a secondary lesion. 

9. The tendency to sarcoma, carcinoma, adenoma, etc., segre- 
gates out and is transmitted as such. 

10. A strong tendency to the location of one or more types of 
tumor in a specific organ or organs, such as the liver, kidney, 
pancreas, mammary gland, etc., is transmitted, owing to the 
segregating out of a peculiar type of tissue in these organs, which 
will respond in a neoplastic or leukemic manner to lesions of any 
kind which furnish a chronic irritation of not too destructive a 
type. 

11. That is, there is a specificity of tissue type, from organ to 
organ in a strain, which will make these organs react in a given 
way to a given type of irritation. 

12. It is, therefore, possible for ancestry to transmit to its 
posterity every possible combination of the neoplastic or leu- 
kemic tendencies which they carry either actually or potentially. 



HEREDITY IN DETEBM1NIN6 TUMOR BiETASTASES 173 

13. This specificity of tissue type in organs, which will insure 
its reacting in a given neoplastic manner or accepting leukemic 
deposits, in response to a given chronic irritation^ is what we find 
to be the nature of cancer heredity, and it is obvious that it will 
make no difference in the nature of the response of such specific 
organ tissue, whether the lesion is primary or secondary in that 
organ or organs. 

Heredity, Iher^ifTe, of a specific type of organ ti^ue, is here 
shown to he the fundamenial influence in determining the incidence 
and location of secondary tumors and of levkemia and pseudoleur 
kemia, just as it is in determining, the incidence and location of 
primary neoplasms. 

It is also pointed out that any apparent testimony of the 
frequent occurrence of secondary tumors in man in tissues where 
primary tumors rarely occur^ would be of no help here, as no 
human strain has ever been even partially analyzed, and no right 
conclusions regarding heredity can be drawn except from analyzed 
strains. 

It is also suggested, that imtil stocks of animals to be used 
in pathological and bacteriological experiments, have been thor- 
oughly tested out as to their inherited potentialities, such 
experiments will be lacking in any adequate control, since 
heredity is not considered although it is tremendously potent. 

REFERENCES 

(1) EwiNo: Neoplastic Diseases, Philadelphia, 1910, p. 86. 

(2) Sltx, Holmes, and Wells: J. Cancer Res., 1919, iy, 207. 

(3) Sltx, HoLia», and Wblls: J. Cancer Res., 1917, ii, 401. 

(4) Sltx, Holmbb, and Wblls: J. Cancer Res., 1921, vi, 57. 

(5) Sltb, Holmbs, and Wblls: J. Cancer Res., 1917, ii, 14. 

(6) LoBB, L: Univ. Penn. Bull., 1906, xix, 113. 

(7) Lbwin: Ztschr. Krebsforsch., 1908, yi, 273. 

(8) Babhfobd, et. al: Third Sci. Report, Imperial Cancer Research Fund, 

1908, 322. 

(9) Woglok: J. Cancer Res., 1918, iii, 47. 

(10) LbCount: Reference in J. Cancer Res., 1917, ii, 6. 

(11) Ewdvg: Neoplastic Diseases, Philadelphia, 1919, p. 143. 



PROCEEDINGS OF THE AMERICAN ASSOCIATION FOR 

CANCER RESEARCH 

FOURTEENTH ANNUAL MEETING 

Held in Cleveland, Ohio, March H, 19il 

1. Rbpobt of teos Council 

The meeting of the Council was held at the Hotel Statler in Cleve- 
landL Ohio, on the evening of Wednesday, March 23, 1921. 

Tne follbwing members were present: Dr. Robert H. Greenough, 
president; Dr. I^ancis C. Wood, and Dr. William H. Woglom. Absent, 
Dr. James B. Murphy, Dr. E. R. LeComit, Dr. Willy Meyer, and Dr. 
James Ewing. As four members of the Council are required for a 
quorum, business was transacted temponuily with the idea of obtaining 
subsequently the vote of another coimcillor. Dr. MUrphy's vote sus- 
taining the action of the Council was subsequently received by mail. 

The treasurer's report was read and accepted. 

The following officers were elected to serve for the ensuing year: 
Dr. James B. Murphy, president; Dr. Wil)y Meyer, vice-president; 
Dr. William H. Woglom, secretary and treasurer (re-elected). 

Dr. H. Gideon Wells was elected Councillor to succeed Dr. E. R. 
LeCoxmt, whose term of office expired. 

The present Council, therefore, with the years of retirement, is as 
follows: 

Dr. F. C. Wood. 1922 Dr. Robert B. Greenough, 1925 

Dr. James B. Murphy, 1923 Dr. Willy Meyer, 1926 

Dr. William H. Woglom, 1924 Dr. James Ewing, 1927 

Dr. H. Gideon Wells, 1928 

The present Editorial Board was continued in office. It is composed 
of 

Editor ^T. Woglom 
ABsociale Editor^ Dr. Prime 
Dr. Bloodgood Dr. Loeb 

Dr. Wells Dr. Ewing 

Dr. TysBer 

The following resignations were accepted : 

Dr. C. L. Alsberg Associate 

Dr. D. B. Phemister Associate 

Dr. Robert G. Leconte Associate 

Dr. F. P. Gay Active 

176 



176 FBOCEEDINGS 

There has been one death during the year — ^Dr. H. H. Janeway, an 
active member, who died on February 1, 1921. 

The following gentlemen were elected to membership: 

Ai^ive 

Dr. Herbert U. Williams, Univ. of Buffalo 

Dr. David Marine. Montefiore Home and Hospital 

Dr. William Ophuls, San Francisco, Calif. 

Dr. H. T. Karsner, Cleveland, Ohio 

Dr. Wm. Bayard Long, New York 

Dr. Leo Buerger, New York 

Dr. James H. Wright, Mass. General Hospital 

Dr. Carl V. Weller, Add. Arbor. Mich. 

Dr. Georgine Luden, Mayo Clinic 

Dr. Henry Albert, Univ. of Iowa 

Dr. Eugene L. Opie, Washington Univ. Sch. of Med. 

Asaociaie 

Dr. John G. Clark, Philadelphia^ Penn. 

Dr. Stuart Graves, Louisville, Ky. 

Dr. Harris P. Mosher, Marblehead, Mass. 

Dr. Kenneth T. Taylor, New York 

Dr. G. R. Minot, Boston 

Dr. Seth Milliken, New York 

Dr. John L. Yates, Milwaukee, Wis. 

Dr. Charles Norris, Chief Med. Examiner, New York 

Dr. D. Crosby Greene, Boston 

Dr. F. S. Mandlebaum, New York 

Dr. Hugh H. Young^ Johns Hopkins Hospital 

Dr. Henry A. Christian, Peter Bent Brigham Hospital 

Dr. Otto V. Huffman, New York 

Dr. L. Duncan Bulklev, New York 

Dr. Joseph A. Blake, New York 

Dr. Ward J. MacNeal, Forest HUls, N. Y. 

Dr. Channing C. Simmons, .'Boston, Mass. 

Dr. Otto Krehbiel, New York 

Dr. Julius Rosenstim, San Francisco, Calif. 

The application of Mr. Donald C. A. Butts, who has not yet pub- 
lished any articles on cancer, was laid on the table, and the application 
of Dr. Rex Duncan was postponed. 

SCIEl^IFIC SESSION 

A letter from Dr. H. J. Conn, Chairman of the Committee on Bacterio- 
logical Technique, of the Society of American Bacteriologists, caUing 
on biologists to cooperate with that Society in helping to secure a reli- 
able domestic source of stains was brought to the attention of the Asso- 
ciation of Cancer Research at its scientific meeting on March 24, by 
the President, Dr. Robert B. Greenough. 

It was moved by Dr. Wood and seconded by Dr. Gaylord that the 
Association encourage the manufacture of American dyes. The motion 
was carried. 



PROCEEDINGS 177 

2. Report on the Journal of Cancer Research 
Dr. F. C. Wood (New York) : 

SUMMARY 

It has seemed to the Council a pity to allow the Journal of Cancer 
Research to lapse, inasmuch as the volumes so far printed contain 
most important papers on cancer research, and when a journal ceases 
pubUcation the libraries put the boimd volumes on the back shelves and 
no one ever sees them again, imless specially inquired for. The Asso- 
ciation has resigned all responsibility for the Journal of Cancer 
Research and the Crocker Fund, Colimibia University, has assumed 
it. There will be no change in form or publishers at the present time. 
The Crocker Fund will have to meet an annual deficit in the cost of 
publication and I hope the members of the Association will do all they 
can to increase oiu* subscription list, because imless this is done the 
Crocker Fund will have to continue the present practice of the Edi- 
torial Board and charge pretty heavily for cuts and r^rints. 

If we could get an additional three or four hundred subscribers it 
would greatly reduce the cost of publication to the individual writers. 

DISCUSSION 

Dr. Robert B. Greenough (Boston) : The Council felt that the Crocker 
Fimd is coming to the rescue of the Association in the matter of the 
Journal of Cancer Research and that there is a distinct obligation 
upon this Association that the Journal should not be abandoned ; for 
much of the work in the past six or seven years in this coimtry has been 
published in this Journal. If abandoned, the material in the earlier 
editions is lost and not readily accessible to new investigators. It is 
thus an obligation to the contributors that the Journal should be con- 
tinued, and apparently only by means of assistance from the Crocker 
Fund can this be done. For these reasons the offer of the Crocker 
Fimd was accepted. 

3. Further Studies in Radiation Dosage 
Dr. F. C. Wood and Dr. Frederick Prime (New York) : 

SUMMARY 

Dr. Wood showed a series of lantern slides illustrating the fact that, 
contrary to statements emanating from Germany, there is no destructive 
carcinoma or sarcoma dose; that is, no fixed amount of x-ray can be 
assumed to destroy the cells of any one tumor, for apparently the dosage 
differs greatly both in man and animals for the same microscopic type 
of tumor. 

As the charts showed, in animal tumors the required dosage is from 
2 to 8 erythema doses, and the sarcomata vary in resistance as well 

THB JOURNAL OF CAKCSR RaSSABCH, VOL. VI, NO. 2 



178 PROCEEDINGS 

^ the carcinomata. One interesting thing is the greatly delayed 
appearance of tumors at a point where the cells received a sublethal 
dose. In a mouse tumor which ordinarily appears within a week and 
gains a fair size in two weeks, the appearance may be delayed for a 
month or more and the tumor grow very slowly afterward. If a trans- 
plantation be made from this tumor into another animal the growth 
rate will rapidly approximate normal. Inasmuch as two or three 
months in a mouse's life is equivalent to the same number of years in 
a human being, the question must be raised whether we are not going 
to see late recurrences, three or four or five years after the sympto- 
matic cure of malignant tumors in man, when such symptomatic cure 
is obtained by either radium or x-ray. 

Most of our experiments have been made by raying the tumor par- 
ticles in vitro and then transplanting them; if the tumor be rayed in 
the mouse the animal is killed in a few days by the radiation. If the 
tiunor be immediately transplanted it grows in the new host, unless it 
has been given a lethal dose. This lethal dose is approximately 10 
to 20 per cent more than when the cells are exposed in vitro. 

DISCUSSION 

Dr. E. T. Bell (Minneapolis): What is your opinion of the treat- 
ment of cancer of the cervix uteri with very high voltage apparatus 
such as is now in use in Germany? 

Dr. F. C. Wood: Repljdng to Dr. Bell's inquiry, whether rays from 
new high voltage machines are more effective than those from the older 
type of apparatus. Dr. Wood said he had not found that there is any 
increase in the destructive action of rajrs given off by tubes running at 
180,000 volts as compared with those running at 120,000 volts, meas- 
uring peak voltages between 12.5 cm. spheres. In all his teste there 
had been no question of absorption or of scattered radiation, the tumor 
having been directly exposed to the rajrs. The advantage of the 
high voltage machines lay in the greater penetration which can be 
obtained, but there is no reason to assume that they would be any more 
effective on superficial growths for equal x-ray dosage. Reports 
received recently from German clinics seem to indicate that some of 
the published results are greatly exaggerated. 

In reply to an inquiry from Dr. Greenough regarding stimulation 
effects, Dr. Wood said that he had not noticed any increase in growth 
rate of the tumors, unless very small doses were given; and that the 
tumors which had been greatly slowed at their first transplantation by 
large doses, grow at their regular rate in the second transplantation. 

In response to an inquiry as to the exact mechanism of the effect 
of x-ray on the cell, and whether the connective tissue does not play 
an important part in the destruction of the tumor cell. Dr. Wood replied 



PROCEEDINGS 179 

that he had not the slightest idea of the nature of the changes causing 
destruction of the cells by x-ray or radium. All we know is that the 
division mechanism is interfered with, and that cells in a resting phase 
are much more resistant to radiation than those in mitosis. Experience 
with animals proves that it takes very large doses to kill all the cells of 
a txmior; those in the center of the growth are killed partially by direct 
action and partially by the thrombosis which occurs early in the smaller 
capillaries. The periphery of the tumor, however, is, as a rule, better 
noiuished and there the cells require maximtun dosage. An absolute 
cure requires the killing of every cell during a sii^e exposure, for if 
divided doses be given, some cells may recover from the radiation 
effects. Dr. Wood had not observed that connective tissue is capable 
of destro3dnjg cancer cells. It is perfectly possible that such cells might 
be inclosed in dense scar tissue and remain quiescent for a considerable 
period, but there is a possibility that they might again begin to grow 
years after their inclusion. This had been frequently observed in human 
tumors in the late recurrences in operation scars. Dr. Wood had seen 
some such recimrences five, six, seven, or eight years after operation. 

In response to an inquiry as to the voltage used, Dr. Wood said that 
this was 120,000 volts, peaJc, measured between 12.5 cm. spheres. The 
tumor fragments used in vitro were about 1 to 2 nun. in diameter. If 
whole tumors be used, or if the tumor be exposed in the mouse, the 
superficial layers nearest the tube will be killed, while the deeper layers 
are not. At this voltage, 10 cm. of tissue absorbed 50 per cent of a>ray 
incident on the surface; this 50 per cent includes not only directly 
transmitted rays but also scattered ones. His own experiments had 
all been devised to avoid the complications brought into the problem 
by scattering and absorption, the effect of which can be easily deter- 
mined by an ionization chamber; this varies in human cases with the 
individual, because of the varying depth of the timiors. Dr. Wood's 
dosage was obtained in a way which made it independent of such 
effects. 

* 

Dr. William Duane (Boston): The difference between the dosage 
measured by Dr. Wood and that reported from Germany, may be due 
to scattering. According to recent reports, at a distance of 10 cm. 
below the sUn scattering may increase the dosage 2) times. If these 
pieces of tumor be surrounded by other bits of tissue, they may perhaps 
receive 2§ times as much radiation as they do under the experimental 
conditions described by Dr. Wood. 

4. High Frequency Z-ray Spectra. 
Dr. WUUam Duane: 

DISCUSSION 

Dr. Wood: We all know, of course, that these experhnents of Dr. 
Duane and the results which he has shown us, form the foundation of 



180 PROCEEDINGS 

all our o^-ray work. They underlie all the measurements of the amoimt 
of the r^-ray which reaches the body ; and when his results on the ph3rsical 
side meet mine on the biological side, we shall know the whole story. 
He is investigating the methods of measurement of the quality and 
quantity of x-ray which comes through our filters, while I am trying to 
determine how much of what he thus has measured is necessary to kill 
a cell. In a few years we shall know the effect of long rays or short. 
The one question which the practitioner has to decide is, whether the 
human body will stand all the radiation necessaiy to kill all the cancer 
cells. If it will not stand this amoimt, we cannot cure cancer with 
radiation; if it will, we can. I recently gave a patient something like 
45 erythema doses for a melanosarcoma df the foot; the blood changes 
were nominal; the only effect was that the patient was somewhat nau- 
seated. The foot was of necessity removed within twenty-four hours 
to avoid absorption effects from the dead tissue. This test shows that 
such a dose can be given on a part of the body where the rays do not 
penetrate important oi^ans; but when it comes to putting large doses 
into the abdominal cavity it is a different question. The question of 
distance is also important. If we could only get a tube such that we 
could place the patient at a distance of a meter, conditions would be 
better than at present, where we work with a distance of 30 or 40 centi- 
meters. The next step is improvement in the wattage capacity of 
the tube. Somebody must design a tube which will stand longer 
quantities of current than we can now employ. The amoimt of energy 
is very small, and exposure, therefore, amounts to six or eight hours, 
which is too much for a patient to stand. If we cannot contrive to 
shorten the exposure by use of a very powerful tube, we shall not g^t 
much farther ahead than we are at the present time. 

Dr. Dtume: The x-ray energy increases as the square of the voltage. 
When the voltage is doubled, four times as much radiation is obtained, 
thus decreasing the length of time required for exposure. 

5. The Occubbence of Squamous-Cell Cabcinoma in Mice 

Miss Maud Slye, Miss Harriet F. Holmes, and Dr. H. Gideon WeUe 
(Chicago): Presented by Dr. H. Gideon Wells: 

StJMBiABT 

Review of the literature on the comparative pathology of carcinoma 
shows that, in general, squamous carcinomas are not so common in 
other a-TiinniLlR as in man, with the possible exception of the dog, in which 
carcinoma of the skin is not infrequent. This infrequency in animalB is 
probably more real than statistics indicate, since the skin carcinoma 
will usually be recognized, whereas the great majority of internal 
growths will escape attention. In horses, carcinoma of squamous 
character is seen occasionally about the external genitalia of both sexes. 



PROCEEDINGS 181 

Fowls and other birds not infrequently show carcinoma of the skin, and 
such growths have also been observed in fish and amphibia. A few 
tumors of this group have been described in rats, and squamous-cell 
tumors of the mouth and skin have been observed in mice as well as a 
few cases of squamous-cell carcinoma arisii^ from the cardiac end of 
the stomach. Most of the squamous-cell carcinomas described in mice 
have occurred in the mammary gland, apparently arising either from 
the ampullae of the lacteal ducts or by metaplasia of a columnar 
cell carcinoma. 

In 28,000 consecutive autopsies performed on mice of the Slye stock, 
which had been permitted to live as long as possible without any 
experimental manipulations whatever, we have observed the following 
instances of primary neoplasms of squamous or stratified epithelial 
structure. 

Primary squamous-cell carcinoma of the skin and mouth 71 

Primary basal-cell carcinoma of the skin 16 

Primary keratinizing cell carcinoma of the manmiary gland 56 

Primary squamous-ceU carcinoma of the stomach 4 

Primary keratinizing cell carcinoma of the lung 1 

Primary squamous-cell carcinoma of the rectum 2 

Primary squamous-cell carcinoma of the vagina 1 

Primary stratified carcinoma of meibomian gland 2 

Primary sebaceous gland adenocarcinoma 1 

153 

The squamous-cell carcinomas of mice are distinguished especially 
by the ii^requency of metastasis and they generally show a relatively 
slight tendency to infiltrate, although we have observed even infiltra- 
tion of the sloill and spinal column. This probably depends on the 
fact that secondary infections usually kill the mice at an early stage 
of the disease. AU the basal-cell carcinomas and the great majority 
of squamous carcinomas arose about the head, neck, and mouth, these 
being the sites of the greatest amount of irritation. Frequently they 
have arisen in healed woimds. 

(A full report of this material will be published in a forthcoming 
number of the Joubnal of Canceb Reseabch.) 

DISCUSSION 

Dr. Wells: In reply to a question regarding the life span of the 
mouse, Dr. Wells rephed that it depends upon the stock. Miss Slye 
has a strain of Japanese waltzing mice which become senile before t^ey 
are a year old. The chief cause of death in old mice is chronic nephritis. 

Miss Slye: The average published life of the mouse is 1^ years, but 
many mice in my laboratory hve to be six years old or more; the average 
age, however, is from three to four years. Every effort is inade to keep 
them alive as long as possible. 



182 PBOCEEDINOS 

Dr. Wells: I may state that these squamous-cell carcinomas in mice 
arise distinctly at a later age on the average than the other carcinomas. 
This is true in human squamous-cell carcinomas, too. 

Dr. WiUiam H. Woglom (New York) : In a paper published in the 
Journal of Cancer Research I have discussed the frequency of 
keratin in manmiary carcinoma of the mouse, and have estimated it at 
about 23 per cent. 

6. The Influence of Heredity in Determining Tumor Metastasis: 
Studies in the Incidence and Inheritability of Spontaneous 

Tumors in Mice. Fifteenth Report 

Miss Maud Slye: 

summary 

Metastases in this stock are somewhat rare. In 29,000 autopsies, 
furnishing something over 4000 primary spontaneous tumors, about 
19 per cent of the growths metastasized. The general testimony of 
those who have discussed metastasis at all, seems to be that the occur- 
rence and location of metastasis is partly a mechanical matter, and 
partly a striking specificity of locaUzation of secondary growths. This 
specificity they make no attempt to explain. 

Ewing, speaking of metastasis in human tumors, states that in 
highly vascular tissues like the lip, stomach, and testicle, very small car- 
cinomas may yield distant metastasis. But these very organs are among 
those whose tumors rarely metastasize in this stock, and then only locally. 
For example, of our carcinomas of the Up none has metastasized; of 
growths in the testis, one metastasized locally; while among carcinomas 
of the stomach three metastasized into the regional lymph-nodes only. 
And this in spite of the close resemblance between these neoplasms and 
human tumors in simflax organs. 

In this stock mammary gland tumors rarely metastasize into the 
regional lymph-nodes, in contrast to those of the human breast. In- 
deed, in many cases in this stock, cardnoma and sarcoma of the mammary 
gland can be seen growing up to the lymph-node but not invading it. 
On the other hand, pulmonary metastases from mammary gland car- 
cinoma and sarcoma are common in this stock. 

These divergences in the metastatic behavior of tumors of similar 
type and in similar organs, require some explanation other than a mere 
mechanical tendency of certain types of tumor to metastasize in certain 
locations. 

Metastasis in this stock has been under observation for ten years and 
the data here given are based on over 29,000 autopsies. 

Briefly stated, these results show conclusively: (1) In any given 
strain the metastatic tumors tend to occur most frequently in exactly 
the same oi^ans in which the primary tumors of that strain occur. 
For example, if a strain is high in primary tumors of the liver, many 



PROCEEDINGS 183 

secondary tumors will be found in that organ. Again, if a strain yields 
large numbers of primary lung tumors, metastasis will tend to occur in 
the lunge from primary tumors in almost any location. A strain high 
in primary kidney tumors is high also in secondary kidney tumors, etc. 

(2) In certain strains there is a tendency for tumors to metastasize 
into certain organs; whereas in other strains, tiunors of the same t3rpe 
in the same organ, pven where they are older and of lai^r growth, 
fail to metastasize into these organs. For example, in certain strains 
practically every carcinoma of the mammary gjiand metastasizes into 
the lungs; I have many strains showing 100 per cent such behavior. 
On the other hand, in other cancer strains exactly this same type of 
mammary gland carcinoma, even where older and of laiger growth, 
never metastasizes into the limgs; and I have many strains showing 
100 per cent of this type of behavior. Indeed, even where tumor emboli 
reach the lungs in these non-pulmonary strains, the emboli f aU to take 
hold and no tumor is formed. This is definite proof that the lung here 
fails to yield, even when tumor cells are present. 

Again, strains which never yield primary lung tumors never yield 
secondary lung tumor. Strain 164 is a striking example of this, one 
family jdelding 36 carcinomas and sarcomas of the mammaiy gland 
without one case of lung metastasis. In this family also there has never 
been one case of primary limg tumor. 

(3) Individuals with secondaiy tumors in any given oi^an seem to be 
as potent as individuals with primary tumors of that organ to transmit 
(by heredity) primary tumors in that same organ. For example, femcJe 
3 with a primary carcinoma-sarcoma of the mammary gland, and sec- 
ondary sarcoma in the kidney, transmitted primary kidney timiors to 
the strains derived from her, one family showing as high as 37 per cent 
of primary kidney tumor. 

Furthermore, individuals with secondary limg tumors seem to be just 
as potent as individuals with primary lung tumors to transmit (by 
heredity) primary tumors of the lui^. Thus, many strains derived 
from a female with a primary manmiary carcinoma with metastasis 
in the lungs, mated with a male showing primary lung carcinoma (double 
lung-cancer parentage) show 100 per cent primary lung carcinoma. 

(4) Etiological meaning: The fact that both primary and secondary 
tumors of a given organ tend to occur in the same strains and fail to 
appear at all in other strains, indicates that heredity is a strong factor 
in determining not only where the primary tumors shaU occur but also 
where the secondary tumors shall occur. 

This fact apparently means that the thing which is transmitted in 
the heredity of cancer is the tendency of an oi^n to yield to cancer, 
whether the lesion is primary in that organ or whether cells from the 
primary growth lodge in that organ and form a secondary lesion. 
Moreover, in strains from which timiors of a certain organ have been 
eliminated by heredity, even where cells from the primary neoplasm 
lodge in such organ they fail to take hold and do not form a secondary 
growth. It is, therefore, evident that not only primary neoplasms, 



184 PBOCEEDINGS 

but secondary tumors also, are determined in their occurrence and 
their location by heredity. 

DISCUSSION 

Dr. G. H. A. Clowes (IndianapoUs) : Has anybody ever attempted 
to determine whether any of these strains are more or less sensitive to 
absence of vitamines? 

Miss Slye: I do not think so. 

Dr. Clowes: I should think that it might be a very important factor 
in proliferation. 

Miss Slye: I have been carrying on some dietary experiments in my 
own laboratory, but they are at too early a stage for me to give any 
data. 

Dr. Wells: Dr. Wood who has investi^ted the development of 
metastases in rats subjected to exploratory incision, in order to deter- 
mine the rate of metastasis after such incision, has found distinct varia- 
tions among control rats of different strains, in respect to metastases. 
These experiments, however, were carried out with grafted tumors, 
whereas Miss Slye's work deals entirely with spontaneous growths. 

Dr. Greenough: As I understand it, all the matings were made before 
it was known that female No. 3 had any tumor. Was her tumor diih 
covered at autopsy? 

Miss Slye: Female 3 had offspring both before and after the appear- 
ance of her tumor. Her mammary tumor was evident some months 
before her death, while the growths of Uver and kidney were found at 
autopsy. Many of these matings are made before the tiunors appear, 
for if one waits for the appearance of a tumor one is apt to lose the 
chance of offspring in a lai^ number of cases. Mice of high cancer 
ancestry shoidd be mated early. In the majority of my cancer mice 
I have records of offspring both before and after the appearance of 
tumors. In the matter of tumor inheritance it makes no difference 
whether the offspring precede by many months the appearance of the 
growth or are bom after its appearance. The grandchild or great- 
grandchild may develop a tumor before the grandmother from whom 
the tumor is inherited. In tumors of the hver and kidney it is difficult 
to make a clinical diagnosis, though this is sometimes possible. Tumors 
of the lung are frequently accompanied by a pecuUar condition of the 
eye, from which it is often possible to make a dinical diagnosis. 

Dr. Clowes: Is there any relation between age and the rapidity at 
which tumors tend to develop and metastasize? 



PROCEEDINGS 



185 



Mi88 Slye: The relation of age and tumor growth has not yet been 
completely worked out, but in general it is true that tumors grow more 
slowly in old mice, as they do in old men. Also it may be said in 
general that the tumors of old mice never grow to any great size, and 
that old mice seldom are able to support more than one tumor. 

7. Pkoofs op the Constitutional Natube of Cancer 
Dr. L. Duncan BuUdey (New York) : 



I. Laboratory 
findings 



II. Statistical 
evidence 



, SUMMARY 

I Cancer not parasitic 
Cancer not contagious 
No cause for cancer 

I Cancer ceU an altered normal cell 
Feeding experiments showing control 
of cancer growth 

'Control of death statistics of cancer and tuber- 
culosis, especially since 1900 

Steady increase of cancer deaths under sm^iy, 
x-rays, and radiimi 



III Bio-chemical f®^^^ ^ cancer; early and late metabolic changes 
evidence I ^^ *^® system before and after the development 
[ of the local cancerous lesion 



IV. Clmical 



Opinion of many celebrated surgeons during the 

last 100 years, to the present time 
Spontaneous cures of cancer reported 
Dozens or hundreds of attested cases of benefit 

or cure of cancer by numerous physicians, in 

this and other coimtries 



DISCUSSION 

Dr. BuUcley: I hope for comments or objections, because if I am 
wrong I want to know it. 

Miss Slye: It might interest Dr. BuUdey and the Association to 
know that there is a standard diet maintained in my laboratory, which 
is identical for tumor strains and non-tiunor strains. With autopsies 
at 29,000 and living inhabitants numbering about 13,000, all fed on 
exactly the same diet, the tumor strains have yielded over 4000 spon- 
taneous tumors, while the non-tumor straiim have never yielded a single 
tumor. No meat is ever used in our laboratory. 



180 PROCEEDINGS 

Dr. W. T. Bavie (Boston): The mortality from tuberculosis in the 
state of Massachusetts is of interest in this connection. A curve 
showing the pn^ressive decrease in mortaUty deviates but little from 
a straight line. Its course does not appear to have been changed by 
any of the activities of the medical profession. Such deviations from a 
straight line as do occur seem to be correlated with influxes of foreign 
people. 

Dr. Wells: I am familiar with Dr. Mayo's paper on the subject of 
the influence of diet on gastric cancer, and it seems to me that Dr. 
Bulkley has mis-interpreted it. Dr. Mayo was discussing the influ- 
ence of local conditions in causing carcinoma of the stomach, and not 
the influence of diet in its constitutional relation. 

8. Relation of Heterosexual Chabactebs to Neoplasms and 

Developmental Erbobs 

Dr. OUo V. Huffman (New York): Read by title: 

SUMMABT 

A preliminary report on the association of heterosexual secondary 
sex characters with neoplasms, especially carcinomata, in patients 
observed by the speaker. In noting the occurrence of status lymphat- 
icus in the clinic he observed that about ten per cent of the patients 
showed some d^ree of heterosexual secondary sex characters, especially 
in regard to the pubic hair and that on the chest and on the face, and in 
regard to the width of the chest and of the hips. He foimd in some of 
these patients such evidence of maldevelopment as a patent omphalo- 
mesenteric duct and gynecomastia. This led him to take note of the 
secondary sex characters of patients with neoplasms. A marked degree 
of heterosexuality in the secondary sex characters is not incompatible 
with normal sexual life and reproduction. Four married patients who 
developed carcinoma showed sUght degrees of heterosexuaUty but the 
marriages were sterile. Another four married patients who developed 
carcinoma and who had some heterosexual characters did not develop 
their neoplasms until after their active sexual life, that is, after the age 
of fifty. The greatest degree of heterosexuaUty was observed in patients 
who developed carcinoma early, that is, prior to fifty years of age. 
This may mean that the degree of heterosexuidity has some etiological 
significance. Between the 100% normal male and the 100% normal 
female we have a whole series of intergrades. Several observers have 
stated that these intergrades are evidently on the increase. Ordinarily 
they are more susceptible than the normal to acute infectious diseases, 
but sanitation and pubUc health measures may be favoring them to 
such an extent that they mature and thereby arrive at the cancer age. 
From the author's observations it would seem that early cancer is rela- 
tively more prevalent among them than among the normal, but he is 
not yet prepared to publish comparative tables that would establish 
this point conclusively. 



PROCEEDINGS 187 

9. End-Result Repobtb of Canceb Operations 
Dr. Robert B. Greenaiigh and Dr. Charming C. Simmons (Boston): 

SUMMARY 

While many reports of the end-results of operations for the cure of 
cancer have been pubUshed, and while certain standards such as the 
arbitrary three or five year limit of time are generally accepted, no 
uniform plan for the reporting of surgical statistics exists, so that it is 
practically impossible to compare the results from one cUnicwith 
those of another. This matter first attracted our attention in 1907| 
when we reported the end-results of operations for cancer of the breast 
at the Massachusetts General Hospital, and again in 1908, when we 
made a study of the end-results of operations of cancer of the mouth, 
tongue, and jaw. In the latter report the following statements were 
made: 

A comparison of these statistics with those of other writers was made and 
a large number of statistical papers were consulted. It was foimd, however, 
that the varying conditions under which the statistics were prepared made a 
comparison difficult without injustice to one or another writer. It tooidd seem 
that a generally accepted standard form for the report of end^reeulta of operatione 
for cancer should he secured. 

In a recent paper on cancer of the breast an arbitrary standard was adopted, 
and that stanaard has been maintained in the present communication. It is, 
briefly, as follows: 

1. A definite period of time has been selected ending; at least three years prior 
to the report, and <Ul of the cases entered in the hospital records under the given 
diagnosis have been investigated. 

2. No case has been accepted as cancer without proof by pathological exam- 
ination, or subsequent recurrence, or autopsy. 

3. Cases which have survived, at last report^ only a portion of the necessary 
three-year period, are eliminated as inconclusive. 

4. Uases not traced at all after discharge from the hospital, and not appearing 
in the mortality statistics of their place of residence, are eliminated as incon- 
clusive. 

6. All cases fulfilling the above requirements are published and counted in 
the statistics with such subdivision into radical and palliative operations as may 
seem expedient. 

It is the hope of the writers that this standard requirement may be adopted 
by others for similar reports. A writer who counts as successful cases those wnich 
have survived a period of less than the arbitrary three-year standard, or who 
accepts the clinical diagnosis of cancer without microscopic examination of the 
specimen, cannot in justice ask to have his statistics accepted for comparison 
with those of surgeons who exercise a more careful scrutiny. It is well said by 
Halsted: "It is especially true of breast cancer that the surgeon interested in 
furnishing the best statistics ma^ in perfectly honorable ways provide them." 

The obvious solution of this difficulty is to publish all cases which fulfill the 
present conditions and allow the reader to draw his own conclusions. If one 
surgeon reports 100 cases, of which 25 are palliative and 75 radical operations, 
his statistics for radical operations may well be expected to be inferior to those 
of the surgeon who performs 50 palliative and 50 radical operations in the same 
total number of cases. Without these figures, however, the opportunity for 
comparison is lost, and the two sets of statistics apparently meet on equal terms. 



188 FROCEEDINQS 

Since the method above described has proved satisfactory in opera- 
tion, the writers would again present it as a standard suitable for adop- 
tion by the profession at large; and as a further example of its value 
they would here report, briefly, a second series of cases of cancer of the 
breast which were operated upon at the Massachusetts General Hospi- 
tal between the dates of August 1, 1911, and April 1, 1914. During 
this period the poHcy of special assignments of cases to different mem- 
bers of the staff was in operation, and to the writers were given all 
cases of diseases of the breast which entered the hospital during that 
time. By following the same plan of reporting adopted in the earlier 
series a direct comparison of the work in the two series of cases is made 
possible. The only discrepancy Ues in the fact that a five year period 
of time was adopted in the second series while the three year period 
was employed, as was customary at that time, in the earlier cases. 

The scheme for the reporting may be suiomarized as follows: 

A. Record all cases entering surgical wards with the specified diagnosis during 

period selected. 

B. Eliminate all re-entries. (No single case should appear twice in the report). 

C. Eliminate all cases recurrent after previous operation in hospital or elsewhere ; 

these are not cases of primary attempt to cure. 

D. Deducting B+0 from A we have the number of cases of cancer available 
for study of operability, mortality, and other operative statistics. These 
cases may then be subdivided as follows: 

E. Cases of radical operation. 

F. Cases of palliative operation. 

0. No operation advisea or performed. 

H. Operative deaths. 

/. Operative mortality H+E-fF. 

/. Operability fradicid operations) E+D. 

K. Operability (all operations) E4-F•^D. 

For the study of end-results of treatment certain cases included in D are 

of no value and should be deducted, vis: 
L. Cases not proved to be cancer either by pathological examination of tissue, 

or recurrence, or autopsy. 
M. Cases untraced after leavmg hospital for required interval of time — three 

years, five years. 
N, Cases that have died of other diseases within the required interval of time, 

and without evidence of recurrence. 
0. The cases remaining after deducting L, M, and N, from D, are available for 

study of end-results, as follows: 
P. Radical operations. 
Q, Palliative operations. 
R. No operation. 

S, Number of cases alive without recurrence (3 years. 5 years). 
T. Number of cases died (over 3 years or over 5 years) without recurrence. 
U, Number of 3 year or 5 year ''cures": all operations. 
F. Number of 3 year or 5 year ''cures": radical operations. 
W, Percentage of 3 year or 6 year "cures": all operations (U+P-f-Q). 
X. Percentage of 3 year or 5 year "cures": radical operations (V+P). 

End-^eavlU: Carcinoma of the hre<ut 

189Jht904 t9U'19t4 

A. Total entries. Carcinoma Breast 613 115 

B. Re-entries (entered more than once) 80 8 

C. Recurrence from previous operation 65 4 



PROCEEDINGS 189 

D. Cases available for study of operability, mor- 

tality, etc 468 103 

E. Radical operation 360 74 

F. Palliative operation 56 20 

G. No operation 52 

H. Operative deaths 15 

I. Operative mortality (H+E+F) 3.6% 

J. OperabUity: Radical operations (E-f-D) 77% 72% 

K. Operability : All operations (E+F-f-D) 89% 01% 

L. Inconclusive cases; Lack pathological exam- 
ination 

M. Inconclusive cases: Untraced 38 5 

N. Inconclusive cases: Died within time limit ... 2 3 

O. Cases available for end-result data 428 95 

P. Radical operations 320 69 

Q. Palliative operations 56 17 

K. No operation 52 9 

S. No. cases alive and well 64 22 

T. No. cases died without recurrence 7 1 

U. No. 5 year "cures" (all operations) 71 23 

V. No. 5 year "cures" (radical operations) 67 22 

W. Percentage of "cures" (all operations) 

(U-i-P+Q) 19% 27% 

X. Percentage of "cures" (radical operations) 

(V-5-P) 21% 32% 

This paper appears in full in the Boston Med. & Surg. Jour. 1921, clxxxv, 253. 

DISCUSSION 

Dr. Bell (Minneapolis) : Very important considerations in the cura- 
bility of carcinoma of the breast are the size of the tumor, regardless of 
its histolc^c type, and the presence of metastases in the axillary nodes 
at the time of the operation. Has Dr. Simmons any information on 
the cured cases relative to these points? 

Dr. Simmons: We divided the cases clinically into four groups. In 
group 1 were placed cases having a small tumor and no palpable axillary 
nodes. Seventy-one per cent of these cases are living without recur- 
rence. Group 2 comprised dightly more advanced cases with small 
palpable axillary nodes. Thirty-three per cent of the cases in this 
group are living. In group 3 were placed the advanced cases, with many 
axillary nodes, in which an attempt at cure by radical operation was 
made. Ten per cent of the cases in this group are living. Group 4 
consisted of advanced cases in which a palliative operation only was 
attempted. live per cent of this group, one case, are living. 

There were two cases of colloid carcinoma in both of which there 
were axillary metastases. These cases are both living without recurrence. 

Infection of the axillary nodes, as shown by the microscopic examina- 
tion, had a distinct bearing on the prognosis of the case. Thus there 
were twenty-four per cent cures in the cases showii^ axillary infection, 
and fiftynsix per cent cures in those where the nodes were not involved. 



190 PROCEEDINGS 



It is interesting to note that in sixty-nine per cent of the cases the 
radical operation did all that could be expected of it; that is, if the 
patient died she died from remote metastases and without recurrence 
in the scar or axilla. 

10. Radixtm in Cancer of the Bladder 
Dr, Oeorge Gilbert Smith (Boston) : 

BmacART 

A report of experiences at the Huntington Memorial Hospital, 
Boston, with the use of radimn in 24 cases of cancer of the bladder. 
Of these cases only 1 might have been suitable for radical removal. 
Seven cases were quoted to show the effect of single applications of 
radium in the effort to determine a standard dosage. Fifteen cases 
were treated by the introduction of screened radium emanation into 
the bladder cavity. Of these 9 cases had a total treatment of 600 mc. 
hours or more. Analysis of these cases showed that while 2 of them 
gave a temporary diminution in the amoimt of tumor and 1 possibly 
showed a total destruction of carcinoma, none was completely cured 
even clinically. Six cases were treated by the introduction of bare 
emanation tubes into the tumor itself. In 3 cases this was done without 
opening the bladder; in 3 it was done by cystotomy. Of these cases 1 
appeared to be completely cured after 1 year. Another case dying six 
weeks after operation showed no carcinoma at autopGfy. With the 
other cases insufficient time has elapsed to determine the insult of the 
treatment. 

Cancbmona 

1. It is useless to attempt to cure with radium infiltrating carcino- 
mata, which involve large portions of the bladder wall. Necrosis of 
the bladder will be brou^t about by any dosage which will materially 
influence the tumor. 

2. Certain superficial cancers of the bladder may be reduced in extent 
by the appUcation of screened radium emanation to their surface. This 
may occur without necrosis of the bladder wall. 

3. To accomplish this effect, 400 mc. hours, with screening of 0.5 mm. 
silver, appHed not of tener than once in six weeks, has been successful, 
and has not caused any considerable reaction in the bladder. 

4. The greatest effect is produced by the first 3 or 4 appUcations of 
radium. 

5. If the tumor begins to grow again, further radiimi applications 
have little deterrent effect. 

6. The best way to employ radiimi in cancer of the bladder is by the 
implantation of bare emanation tubes in the tumor, allowing one tube 
to each cubic centimeter. Steel needles containing radiimi may be 
employed in the same way, except that they must be withdrawn after 
adequate exposure has been made. 



PBOCBEDINOS 191 

7. The necrosis caused by the implantation of radium in bladder 
tumors persists for at least three months. 

DISCUSSION 

Dr. Wood: I think that these papers by Dr. Smith and Dr. Simmons 
are admirable. It is important that we laboratory men check up our 
results on human beings ; we care nothing for rats and mice in themselves. 
One interesting point is the statement that a radiated tumor is resistant 
after scar tissue has been formed. . The observation has been made by 
others also. I can not think thai the carcinoma cell undei^es any 
change/ simply because of the presence of scar tissue. The radium 
dosage for cancer cells in tissues not highly vascular, or in contact with 
normal tissue, is very much higher than for the same cancer cells where 
thrombosis can take place and cause extensive destruction. For mouse 
cancer cells, the dose is 2400 mc. hours at a distance of two centimeters; 
you see how resistant the cell is. Now Dr. Simmons has shown that 
630 mc. hours are all that can be allowed, so that we are still far from 
able to guarantee killing the cell, or from promising that other cells will 
in any way affect the life of the cancer cell. I do not believe, for exam- 
ple, that lymphocytes or connective tissue cells destroy cancer cells; 
there is not the slightest evidence to that effect. Indeed, the cancer 
cell may he quiescent in the connective tissue for seven, ten, or fifteen 
years. 

11. Pboblems in Cancer Research 
Dr. Montrose Burrows (Saint Louis) : 

(Published in the Joxtrnal of Cancer Research, 1921, VI, 131) 

discussion 

Dr. Burrows: In reference to the discussion on the alkalinity of the 
blood in cancer patients, I wish to state that we have found no direct cor- 
relation between the alkalosis in many cancer patients and the anaemia 
which is invariably present. Alkalosis in anaemic patients is a weU 
estabUshed fact. So far our conclusions in relation to anaemia have 
been drawn from a comparison of the blood tests for pH which we have 
had made and the blood picture recorded on the hospital history. We' 
are now anxious to make all these tests for anaemia ourselves. We 
also feel sure that other conditions such as lesions of the pancreas and 
kidneys will alter materially the pH determinations of the blood in 
the cancer patients. Certain of our cases as well as others cited in the 
literature have indicated this fact. 

In referrinjg to Dr. Clowes's statement about a lipoid membrane, I 
mBt to call his attention to the fact that such exists in the tissue culture. 
We noted this fact in the Uterature in an article on "TTie cultivation of 
bladder and prostatic tumors" (Jour. Urol., vol. 1, no. 1^. In analyzing 
the property of stereotropism introduced for body cells by Harrison, 



192 PBOCEEDINOS 

we found that these cells were not in contact with the cover glass or 
free surface of the medium as Harrison stated, but separated from the 
surfaces by a surface film of substance which is insoluble in the medium. 
This was particularly easy to demonstrate in the cultures of bladder 
and prostatic carcinomata. 

Returning to the question of blood alkalosis in cancer, our work has 
further suggested to us that the alkalosis is due to the addition of an 
alkaline substance from the tumor into the blood. The pH becomes 
changed because of the lack of a very active compensating mechanism 
for alkaUes (BayUss). 

The demonstration of an alkalosis in the plant cancers and in the 
culture of the Bacillus tumefaciens gives further confirmation of this 
fact. 

12. Massage and Metastasis 

Dr. L. C. Knox (New York, ty invitation) : Presented by Dr. F. C. 
Wood; 

SUMMABY 

The general conditions underl3ring the formation of metastatic 
deposits from maUgnant tumors have long been of interest to the 
sui^on as determining the scope and direction of his operative pro- 
cedures and even the feasibiUty of operation. 

In this country, although the dermatologists have always made 
biopsies in doubtfid cases, there has been during recent years a great 
deal of discussion among surgeons as to the danger of incising timiors. 
Isolated instances of apparent distribution following diagnostic inci- 
sions have led to the widespread f eehng that such incisions tend to dis- 
tribute the tmnor. This attitude, strangely enough, has existed only 
in the United States; the English, French, and German sui^ons do not 
seem to have considered the matter as of importance. 

On the other hand, the relationship of massage to the production 
of tumors has excited but Uttle interest in the minds of the practitioners 
of surgery and there are but few recorded clinical observations of the 
spreading of tumors by massage, though such distribution takes place 
very frequently under the administration of the mechanical treatment 
used by the osteopath and the chiropractor to "disperse" timiors. 
Some of the most extraordinary instances of widespread metastatic 
involvement of the entire body have been seen following massage of a 
carcinoma of the breast. Several instances, in which very extensive 
and early metastasis occurred after small tumors had been repeatedly 
examined and rather firmly squeezed or handled by the physician mak- 
ing the examination, called the writer's attention to the importance of 
this phase of the metastasis question. 

Though the problem is one easily attacked from an experimental 
point of view, but Uttle work has been done on it, with the exception 
of experimentia by Dr. E. E. Tyzzer, who some years ago observed that 
massage of a mouse tumor considerably increased the number of metas- 



PROCEEDINGS 193 

tases. His experiments were confined to only one type of tumor; and, 
therefore, it seemed of great practical and scientific interest to extend, 
if possible, his observations to a large variety of tmnors of different 
histological forms so as to correlate the results obtained from animals 
with those observed in human beings. 

It is obvious that a small-cell tiunor in which the cells are closely 
related to the blood-vessels, such as the lymphosarcomata, would metas- 
tasize early and extensively, while the firmer fibrosarcomata might be 
expected to offer considerable resistance to the removal of tumor par- 
ticles into the l3maph-channels and the blood-vessels. As the lym- 
phatic cfystem in animals is not as extensive as that in man, metastasis 
is most frequent by way of the blood-vessels; hence, the tumor cells 
reach the lui^ first, and the effect of massage would be expected to 
increase very greatly the number of secondary tumors in these organs. 
This was found to be the case. 

Two series of experiments were carried out with the same technique, 
one two years after the other. The first consisted in the inoculation of 
mice and rats with six strains of carcinomata and four strains of sar- 
comata. About 500 animals survived for the completion of the experi- 
ment. As soon as the tumors were palpable, half of them were massaged 
for thirty seconds on alternate days for about two weeks; the other half 
were used for controls. After this, the tumors were removed by opera- 
tion, and the animids were killed at the end of thirty days. The lungs 
of these animals, as well as those of the animals which died spon- 
taneously during the course of the experiment, were examined micro- 
scopically, and the number of metastases and emboli was noted. In 
the second experiment only one strain, a polyhedral-cell sarcoma, was 
used with 50 mice. The results of the two experiments were approx- 
imately the same. 

It was found that in all but three strains there was a lai^r percentage 
of lung emboU in the massaged mice than in the controls; that there 
was also usually a higher percentage of actively growing metastases; 
and that the total number of metastatic particles of both quiescent 
emboli and growing metastases constantly increased, the variation 
being from 1 per cent to 37 per cent. In this respect, little difference 
was found between the polyhedral-cell sarcomata and the carcinomata. 
It was found, also, that fibrosarcomata are not so apt to metasta- 
size when manipulated as are the other types. 

13. FuBTHEB Experiments on Sensitization to Heat by Radiation 
Dr. WiJUam T, Bavie (Boston): 

SUMMARY 

The experiments reported in this paper concern the changes which 
take place between the time of radiation and the appearance of the first 
visible effects. 



194 PBOCEEDINGS 

At a previous meeting I have discussed the rate of recovery of 
Paramecium from the destructive action of fluorite rays, and the sen- 
sitization to heat resulting from the exposure to fluorite rays. 

My interpretation of the results of these investigations was incorrect, 
because through faulty methods of experimentation the effects of ozone 
formed by the Ught were not excluded. 

The experiments reported in the present communication were con- 
ducted in such a manner as to exclude the effects of ozone. 

A very definite sensitization to heat was demonstrated. 

By increasing the temperature of the oi^nism after the radiation, 
the length of the latent period is shortened. All of the changes pro- 
duced by the radiation are intensified and the total nimiber of deaths 
for a constant dosage is increased. 

The effect is greater the higher the temperature and the longer the 
time during which the organism is maintained at the increased tem- 
perature. 

The temperatures used were not high enough to affect normal 
unradiated organisms, nor were the effects to be observed if the organ- 
isms were subjected to the increased temperature before, the exposure 
to the rays. 

DISCUSSION 

Dr. Clowes: At what temperature do these changes start? 
Dr. Bovie: Room temperature, 22®C. 

Dr. Clowes: You must take into consideration variation in tem- 
perature, for radiation is ultimately chemical. 

Dr. Bovie: We have a thermo-couple connected with the slide, and 
make certain that the temperature is back at normal before we radiate. 
It would not make a great difference, however, because light reactions 
are not influenced by changes in temperature. 

Dr. Wood: It may interest Dr. Bovie to know that we are not able 
to detect any difference in the cancer cells between heating first and 
radiating afterward, or radiating first and heating afterward. The 
point we used was the death point determined by animal inoculation. 
This is far from what he is working with. We can observe it accurately. 
We get an approximation to a logarithmic curve but whether this is 
the true form of the curve is doubtful. 

14. Malignant Tumors of the Thyroid 
Dr. Louis B. Wilson (Rochester, Minn.) : 

SUMMARY 

This paper presents an analysis of the pathological data concerning 
290 patients with malignant tumors of the thyroid examined in the 



PROCEEDINGS 195 

Mayo Clinic between January 1, 1901, and January 1, 1921. Lantern 
slides illustrating the various histologic types of tumor were shown 
and the clinical course of the disease in cases of the various types dis- 
cussed in connection therewith. The following is a summary of the 
principal points in the paper: 

1. Malignant tumors of the thyroid are much more frequent than is 
generally beUeved. Correct cUnical dia^osis is frequently missed, 
(a) because they may have periods of development of from five to 
fifteen years and patients are not followed up long enough after opera- 
tion; and (b) because not infrequently the tumor is relatively small and 
the character of metastasis is not determined, owing to the rarity of 
necropsies. 

2. Pathologic diagnosis is difficult owing to the great variation in 
the histolc^y of the tumor and its resemblance to that of non-malignant 
processes. 

3. There has been a marked failure of American surgeons to report 
their cases of malignant tumors of the thyroid; this should be corrected. 

4. Sufficient observations are not at hand for determining the 
ge(^aphic incidence. 

5. The age incidence at the date of diagnosis is greatest in the fifth 
decade. 

6. The distribution by sex is about one man to two women. 

7. Patients usually seek medical advice on the occasion of recent 
rapid growth in a long standing nodular tumor of the thyroid. Some 
give histories of slow continuous growth. 

8. Early thorough operation gives a fair percentage of cures. Pal- 
Uative operation in late cases with extensive local involvement is 
warranted. 

9. Patholc^c diagnosis must take into account the usual ilevelop- 
ment of malignant tumors of the thyroid from proUferating embryonic 
adenomas. 

10. The pathologist must be thoroughly familiar with the charac- 
teristics of proliferating adenomas (as first described by Langhans) in 
all their stages. 

11. The pathologist must be on the lookout for a possible relationship 
between bizarre metastatic growths and tumors of the thyroid. 

12. The pathologist, in his diagnosis for the guidance of the surgeon, 
must consider the relative preponderance of proliferative and degenera- 
tive processes in the tumor; but a proUferating adenoma in a patient 
of cancer age should not be considered benign imless the process of 
degeneration is very extensive and thoroughly overbalances that of 
proliferation. 

DISCUSSION 

Dr. Wells: My experience in the Chics^o district corroborates Dr. 
Wilson's statement; there, also, tumors of the thyroid are not imcom- 
mon. It is hard to find a normal thyroid in experimental dogs in 



196 PROCEEDINGS 

Chicago, and I have many sent to me with carcinoma of the thyroid. 
There have been but a few definite tumors of the thyroid in mice in the 
Slye stock. 

15. EXIJERIMENTAL PrODUCTIGN OF TUMORS 

Dr. F. C. Wood: 

SUMMARY 

Dr. Wood reported the experimental production of tiunors in the 
Crocker Laboratory by Dr. Frederick BuUock and Miss M. R. Curtis. 
The method employed was to feed rats with ova from cat feces. The 
ova were those of the Tenia crassicollis, a frequent inhabitant of the cat 
intestine. 

The oi^anism penetrates the mucous membrane of the intestine, 
passes to the Uver and there forms a cyst. In the wall of such cysts, 
sin^e or multiple sarcomata develop after a period which is never less 
than eight months. These sarcomata are of two main types, spindle- 
ceU and large polyhedral-cell. They are highly malignant, metas- 
tasizing throughout the animals, and are transplantable, giving a high 
percentage of successful transplants at the first inoculation. 

The importance of the discovery is the large jdeld of tumors and ease 
of the method, in contrast with the laborious tar painting process. 
One strain of animals gave approximately 50 per cent of tumors among 
all rats infected. 

DISCUSSION 

Dr. BeU: Has Dr. Wood tried any filtration experiments? 

Dr, Wood: Some years ago large series of filtrations were carried out 
on various tumors in the Crocker Laboratory but without success. 

Dr. BeU: I woidd Uke to ask whether the blood shows any changes in 
these very maUgnant sarcomata? Does it surest a leukemia? 

Dr. Wood: No changes at all. 



FURTHER INVESTIGATIONS ON THE ORIGIN OF 

TUMORS IN MICE 

Vn. TUMOR AGE AND TUMOR INCIDENCE 

LEOLOEB 

From ihi$ Department of Comparative Pathology of Waehinffton Univereity School 

of Medicine, 8t, Louie 

Reoeived for publication November 25, 1021 

1. In our former publications we have analysed the relation 
between tumor age and tumor rate in the various strains of 
mice (1). We found that in those strains in which the tumor 
rate was high the tumors tended to appear at an earlier period 
of life than in strains in which the tumor rate was lower. In 
addition, we found indications that there existed a special tumor 
age in certain strains of mice. In general cancer age was trans- 
mitted by heredity as well as cancer rate. 

The present conmiunication differs from our preceding ones 
in the following respects: (a) We analyse here our total material 
connectedly, while previously we considered only certain parts, 
(b) The method of computation of the relations between tumor 
age and tumor rate, which we used formerly, was complicated 
and made a comparison between different strains and groups of 
mice difficult. We now make use of a simpler way of figuring 
out the relation between txunor age and tumor rate and thus 
are enabled to compare more readily the different strains, (c) We 
extend our analysis in various directions and explain our results 
more fuUy on the basis of multiple factors. 

2. In order to determine the tumor age of a certain family 
strain or group of strains we proceeded in the following manner. 
We determined the number of mice alive in the beginning of each 
of the three age periods (I age period, 7-12 months; II age 
period, 13-17 months; III age period, 18 months and older.) We 

m 

TBB JOITBVAL Of CAHCSB RMBASCS, VOL. YI, NO. S 



198 LEO LOEB 

then determined the number of mice which developed tumors 
m each of these age periods and figured out what percentage 
of mice were in each age period affected by cancer. These per- 
centage figures we then reduced to a basis of 100 per cent cancer 
incidence, in order to obtain figures for the cancer age in dif- 
ferent strains which were independent of the absolute number of 
tumor mice in each strain. We may give as an example of this 
mode of determination the figures obtained for the total of all 
81 + 328 strains; the last figure on each line indicates the num- 
ber of mice in which tumors developed in each age period in a 
unit of 100 tumor mice. The figure 56.4 per cent indicates the 
tumor incidence of the whole strain. 

8} + 328 

I age period. 566 mice 194 tumor mice » 34% 27% 

II age period. 206 mice 117 tumor mice *- 66% 44% 56.4% 

III age period. 24 mice 9 tumor mice » 38% 29% 

3. We divide all our strains into high, medium, and low tumor 
rate strains. The high tumor rate strains have an incidence of 
more than 40 per cent. The medium tumor rate strains have 
an incidence varying between 20 and 40 per cent, and the low 
tiunor rate strains have an incidence below 20 per cent. 

In our lists we state first the name of the strain, then we give 
the number of mice and tumor rate in each strain and in the 
last three rows the tumor incidence in each age period, reduced 
to a unit of 100 tiunor mice. Added in brackets to these figures 
are in certain strains the absolute percentage figures of tumors 
in each age period. Wherever the name of the stram as a whole 
is in brackets, the figures were merely given in the list; they were 
for obvious reasons not used in figuring out the incidence in 
each age period in all of the strains belonging to one of the classes 
(high, medium, low rate tumor mice). 

If we compare the figures giving the percentage of mice affected 
by cancer in the three age periods, we may draw the following 
conclusions: (1) If the number of mice in a group is small, for 
instance, below 50, or even somewhat higher, the percentage 
age figures are no longer of value; if the number of mice exceeds 
one hundred, the percentage figures begin to be much more trust- 



OBIOIN OF TDMOBS IN MICB 



199 



High rcUe tumor mice 



ITBAIN 



(London Blue and White) 31 

Total 8§ + 328 566 

(8J + 328A) 118 

(8J + 328B) 226 

[8 J + 328 (new cross)] 146 

[782a (- 8J + 328)] 76 

(English) 24 

(EnglishA).. 175 

[Family 101 (English)] 99 

(English Sable A) 252 

(English Sable B) 48 

[Family 437 (English Sable)]. . . 55 

(Smaller English Family) 36 

[344 + 328 and (344 + 328) + 

437] English Sable 34 

TotalEnglish 689 

[(European 151 + Id of No. 10) 

+ 101 (English) Total] 62 

[(European 151 +Id of No. 10) 

+ 101 English Al 27 

" (EnglishB) 35 

Total Michigan Wild + 

EnglishlOl 70 



BBBOF 
MJCB 



(English A) 

(EnglishB) 

■European 151 + Id of No. 10 

(November 3)1 

Total European 151 + lid of 

No. 10 (Novembet 8) 

(Total European A) 

(Total European B) 

(8i + IIdofNo. 10) A 

English Sable + [(European 

+ 103) Fi + Illd of No. 10] 

F, 

European + English Tan 

(daughter of tumor mouse 

146 B) 

(Silver + 10 B).. 

Total English Tan + German. 
(English Tan + German A) . . . 
(English Tan + German B). . . 



50 
20 

254 

187 
96 
91 

82 



6 



33 
11 
60 

44 
21 



TUMOB 

BATB 



per cent 

55.0 
56.4 
51.0 
56.0 
57.5 
60.0 
46.0 
63.0 
71.0 
70.0 
67.0 
82.0 
56.0 

79.0 
67.6 

42.0 

55.0 
31.0 

48.0 
58.0 
25.0 

72.0 

54.5 
65.0 
44.0 
49.0 



83.0 



42.0 
64.0 
46.6 
49.0 
43.0 



PBBCBMTAGB OP TUMOBB XK 



I ace period 



percent 

18.0(26.0) 

27.0(34.0) 

29.0(84.0) 

25.4(32.0) 

26.5 

35.0(38.0) 

25.0(29.0) 

20.0(29.0) 

21.0(45.5) 

28.0(46.0) 

20.0(27.0) 

37.0(58.0) 

46.0(42.0) 

46.0 
25.0^40.6) 

25.0(26.0) 

40.0 
17.7 

21.0(23.0) 

19.0 

29.5 

15.0(25.0) 

15.0 

14.0 

17.5 

9.0 



0.0 



4.6 
17.0 
24.0 
21.0 
33.0 



II ace period 



per cent 

37.0(54.0) 

44.0(56.0) 

42.6(50.0) 

41.0(52.5) 

44.0 

65.0(70.0) 

32.0(37.0) 

45.0(64.0) 

32.0(09.5) 

34.0(54.5) 

49.0(65.0) 

63.0(100) 

54.0(50.0) 

54.0 
39.0(62.5) 

36.0(37.0) 

60.0 
24.0 

28.0(31.0) 

29.0 

21.5 

33.0(55.0) 

36.0 
36.0 
33.5 
35.0 



38.0 



57.5 
20.0 
41.0 
35.0 
67.0 



III ace period 



per cent 

45.0(66.0) 
29.0(38.0) 
28.4(33.3) 
33.6(42.0) 
29.5 

0.0 (0) 
43.0(50.0) 
35.0(50.0) 
47.0(100) 
38.0(62.0) 
31.0(40.0) 

0.0 

0.0 

0.0 
36.0(54.5) 

39.0(40.0) 

0.0 
58.3 

51.0(57.0) 

52.0 

49.0 

52.0(88.0) 

49.0 
50.0 
49.0 
56.0 



62.0 



37.9 
63.0 
35.0 
44.0 
0.0 



200 



LEO LOEB 



Unknown (probably English) 

Total 108 English + (8 + 
German) 

|188 English + (8 + German 
A)l 

|108 English + (8 + Gernum 
B)l 

(German A) 

(8 + German A) 

IFamily 240 of (8 + German)] 

Combined (SOver + English A) 
and SUver + (English Sable 
C) 

English Sable 844 + German. . 

English Sable + (English 
Silver + Id of No. 10) 

English Sable + (European + 
Id No. 10) 

(English Sable + Cream Y) . . 

603 (- English Tan + Cream) 
+ 773("8J + 328) 

794 (German + Carter) + (8 + 
German) 

794 (German + Carter) + [198 
(English) + (No. 8 + Ger- 
man) F|1 

Waltzer + No. 8 

240 (8 + German) + Cream. . . 

8i + English Sable 

Waltser + English Orange and 
combined cross 



inni- 

BBBOr 



32 

217 

82 

135 

20 

244 

14 

40 
45 

23 

125 
68 

35 

90 



37 
52 
38 

48 

65 



TUMOB 



fW cent 

72.0 

54.0 

63.0 

49.0 
50.0 
41.0 
43.0 



57.5 
64.5 

65.0 

55.0 
53.0 

60.0 

49.0 



61.0 
46.0 
76.0 
61.0 

65.0 



31.0 

18.5 
12.0 
18.0 
29.0 



PUKSDITAOS OV TUWMM HI 



laci period 



per cent 

30.0 



per eeni 

136.0 



20.0(28.0) 41.0(58.0) 



23.0 
20.5 

23.0 

25.0 
14.0 

38.0 

18.0 



21.0 
20.5 
11.0 
40.6 

36.0 



Ilaca period 



69.0 

35.0 
36.0 
40.0 
33.0 



67.0 
26.0 

77.0 

40.0 
46.0 

62.0 

25.0 



26.0 
31.5 
37.0 
59.0 

64.0 



mage period 
per eent 

34.0 

39.0(56.0) 

0.0 

46.5 
52.0 
47.0 
0.0 



20.0 
53.5 

0.0 

35.0 
40.0 

0.0 

57.0 



53.0 

48.0 

52.0 

0.0 

0.0 



worthy. (2) With this restriction as to numbers, we may con- 
clude that the age distribution of tumors is at least as charac- 
teristic of strains and serviceable for their distinction as the 
tumor incidence; in fact in certain cases it is a finer instrument 
for the characterization of strains than the tumor rate. The 
tiunor rate of two strains may be similar, but the tumor age 
may allow a differentiation between the two strains. 



OBIGIN OF TUMORS IK MICE 201 

Thus if we consider especially the high tumor rate strains 
we notice that the tumor rate of European 151 and I daughter 
of No. 10 and of European 151 and II daughter of No. 10 on the 
one hand and of 8^ + 328 and English and its substrains on the 
other hand do not differ very much, while the tumor age is quite 
distinct. 

The percentage figures for the first and third age period are 
similar in the former two strains and differ very much from the 
figures in the latter strains. There is even a difference between 
English and 8^ + 328, both very high rate txmior strains; in the 
English the tumors appear somewhat later than in the 8| + 
328 strain. This comes out even more clearly if we consider 
the percentage figures for the third age period than those for 
the first age period. 

4. If we compare the tumor ages of the substrains with those 
of the main strains we find that on the whole they .agree provided 
the figures used are not too small. This applies for instance to 
the English, 8^ + 328 and European + I or II daughter of 
No. 10 and their substrains. 

In the case of the strain ^'unknown'' there was some indica* 
tion that it was a substrain of the English. Not only the tumor 
rate, but also the tumor age of this strain furnish confirmatory 
evidence for this conclusion. 

5. In crosses between English and the majority of other strains 
in which the tmnor incidence is high the tumor age is similar to 
that of the English. 8} + 328 is a cross of the female offspring 
of a particular English timaor mouse with an 8^ male. Here we 
have, owing to the influence of the English mother, a very early 
tumor age which exceeded even the average age of the English 
mice. On the whole, we find the English or a similar tumor age 
in the following additional crosses: (European 151 + I daugh- 
ter of No. 10) + 101 English, Michigan Wild + English 101. 
(European 151+ I daughter of No. 10) + 101 English) A. 
English Tan + German, 198 English + (8 + German), Silver 
+ English, English Sable + German En^h Sable + (English 
Silver + I daughter of No. 10), English Sable + (European + 
I daughter of No. 10) (English Tan + Cream) + (8J + 328), 



202 



LEO LOEB 



Medium rate tumor mice 



BTRAJK 



(Complete English-Cream Hy- 
brids) 

English-Cream Hybrids, medi 
um rate 

(English Sable 344 + Black 
Cream) 

(Black Cream + English 
White) 

(Cream + English, October, 
1913) 

(White Cream + White Eng- 
lish) 

[Cream-English (Descendants 
of English Sable 1031)] 

(English Sable + Cream Y, 
high rate) 

(693 + Cream) 

(Total English Tan (121) + 
Cream] 

Total London 

(London A) 

(London B) 

(London C) 

(London 481) 

(London Blue and White) 

Total European + 8 Fb 

(European + 8 F»A) 

(European + 8 F»B) 

Total Heitler 

(HeitlerA) 

(Heitler B) 

101 (English) + (European + 
103) 

Total European + English Tan 

(European + English Tan A) . . 

(European + English Tan B). . 

(June, 1914, Family of Cream 
B) 

TotalNo.8 

(N0.8A) 

N0.8A1 

(No.8At) 

(N0.8B) 

Carter 



NITM- 

BBBOr 

IIICB 



1057 

739 

92 

143 

109 

27 

26 

68 

7 

267 
452 
120 

61 
197 

43 

31 
151 
125 

26 
196 
102 

94 

152 

109 

76 

33 

40 

243 

213 

145 

68 

30 

67 



TUMOB 

BATB 



per cent 

24.3 

323 

23.0 

39.0 

23.0 

29.0 

31.0 

53.0 
28.5 

31.0 
28.0 
27.0 
38.0 
28.0 
0.0 
55.0 
28.0 
30.0 
23.0 
25.0 
27.0 
22.3 

34.0 
32.1 
28.0 
42.0 

35.0 
27.5 
30.0 
34.0 
22.0 
10.0 
39.0 



PBBCBNTAOB OW TUIIOB8 Ilf 



I ace period II ace period 



per cent 

20.0 (8.6) 
19.0(11.3) 
18.5 
19.0 
12.5 
36.0 
6.9 
14.0 



22.0(12.0) 
22.0(10.9) 
13.0 (7.6) 
29.0(16.0) 
11.0(25.5) 

18.0 

12.5 (6.0) 
12.5 

9.5 
12.5(61.0) 

8.5 (4.0) 
19.0 (8.5) 

5.5 (4.0) 
7.4 (4.0) 
4.0 
4.6 

4,6 

12.0 (7.0) 

11.4 (7.1) 
17.0(11.0) 

0.0 (0.0) 

0.0 (0.0) 
26.0 



per c4nt 

42.0(18.0) 

41.0(24.4) 

40.0 

47.0 

31.0 

26.0 

44.6 

46.0 



38.0(21.0) 
45.0(22.0) 
36.0(19.0) 
56.5(32.0) 
49.0(21.0) 

37.0 

44.25(21) 

48.0 

26.0 

64.0(31.0) 

91.5(43.0) 

39.0(18.0) 

39.5(29.0) 
50.0(27.0) 
45.0 
57.5 

33.5 

27.5(16.0) 
32.2(19.0) 
33.0(21.5) 
24.5(15.4) 
0.0 (0.0) 
32.0 



III ace period 



per cent 

38.0(16.1) 

40.0(24.0) 

41.5 

34.0 

56.6 

38.0 

48.6 

40.0 



40.0(22.0) 
33.0(16.0) 
61.0(27.0) 
15.8 (8.0) 
25.5(11.0) 

46.0 

44.25(21) 
39.5 
64.5 

23.5(11.5) 
0.0 (0.0) 
42.0(19.0) 

66.0(40.0) 
42.6(23.0) 
61.0 
37.9 

61.9 

60.6(36.0) 

56.4(36.0) 

60.0(32.0) 

76.6(47.3) 

100(20.0) 

42.0 



ORIGIN OF TUMORS IN MICK 



203 



■TBAZlf 



Total 8i + II daughter of No. 

10 

(8i + II daughter of No. 10 

A) 

(8} + II daughter of No. 10 

B) 

(8J + II daughter of No. 10) + 

lid of No. 10 

(European 151 + Id of No. 10} 

+ 101 English B 

Total Silver + Id of No. 10. . . . 

(Silver + 10 A) 

(SUver + lOB) 

Total medium rate Cream + Id 

of No. 10 

(Total Cream + 10 includes one 

low rate cross) 

(Cream + 10 A) 

(Cream + 10 B) 

(White Cream + Id of No. 10) 
(Michigan Wild + English 101 

B) 

Total German 

(German A) 

(German B) 

Totals -f German 

(8 + German A) 

(8 + German B) 

(Family 240 = 8 + German). . . 

German + Carter B 

794 = German + Carter 

English (344) + 8J F4 and F«. . 
Cream + [198 (English) -f (8 

+ German) F4I 

Black Cream + European 

European + Cream 

(Waltzer + White English) + 

(Cream + 10) 

Vermont Wild + English F,. . . 
(f Waltzer J Cream) + No. 8. . 

German + No. 6 F4 

Waltzer + White English and 

(Waltzer + White English) 

+ English 

(Cream B June 1914 Family).. . 



NUM- 
BER OF 
IftCB 



110 

82 
28 
72 

35 
245 
234 

11 

370 

498 

174 

47 

149 

20 

42 

20 

22 

373 

244 

129 

14 

113 

25 

155 

33 

104 
25 

19 

373 

32 

52 



99 
40 



TUMOR 
RATS 



per cent 

40.0 

49.0 

14.0 

36.0 

31.0 
37.0 
36.0 
64.0 

34.0 

27.6 
36.0 
26.0 
34.0 

25.0 
40.5 
50.0 
32.0 
34.0 
41.0 
21.0 
43.0 
33.5 
36.0 
25.0 
33.5 

32.0 
28.0 

21.0 
36.3 
31.0 
37.0 



26.0 
35.0 



PERCENTAOB OF TUMORS IK 



I age period 



per cent 

8.5 

9.0 

0.0 

8.9 

17.7 
17.0 
17.0 
17.0 

10.0 (5.4) 



8.4 

5.5 

22.0 

12.0 



(8.0) 



29.5 

15.5 a9.0) 
12.0 
23.0 

14.0 (8.0) 
13.0 
23.0 
47.0 
22.0 
5.0 
21.0 
36.5(15.0) 

11.0 
5.3 

18.9 
11.0 
28.6 
11.9 



29.0 
4.6 



II age period 



per cent 

40.0 

35.0 

100 

20.8 

24.0 
27.0 
26.0 
20.0 

32.0(17.0) 

29.0 
28.0 
46.0 
33.0(20.0) 

21.5 

30.3(37.0) 

36.0 

51.0 

42.0(24.0) 

40.0 

51.0 

53.0 

31.0 

29.0 

39.5 

73.5(26.0) 

32.0 
49.0 

81.1 
30.0 
46.4 
24.0 



28.0 
33.5 



III age period 



per cent 

51.5 

56.0 

0.0 

70.3 

58.3 
56.0 
57.0 
63.0 

58.0(31.0) 

62.6 
66.5 
32.0 
55.0(38.0) 

49.0 

54.2(66.0) 

52.0 

0.0 
44.0(25.0) 
47.0 
26.0 

0.0 
47.0 
66.0 
39.5 

0.0 (0.0) 

57.0 
65.7 

0.0 
59.0 
25.0 
64.1 



43.0 
61.9 



204 LEO LOSB 

German + Carter + (198 (English) + (No. 8 + German) F,), 
8^ + English Sable, Waltzer + English Orange. Waltzer No. 
8 is similar. It is different, however, in a cross between Eng- 
lish and Cream which approached the tumor rate of the English, 
namely, English Sable + Cream Y. The tumor rate is here 
55 per cent, but the tumor age is distinctly different; it is in- 
termediate between that of the English and Cream. 

German, 8 + German, German + Carter have a later tumor 
age; the same applies to the crosses between these strains. Some- 
what later even is the tumor age of (8 + German) + Cream, 
owing to the influence of the. Cream; yet the influence of the 
8 + German causes the tumor age to be earlier than that of the 
Cream. In a similar way the tumor age of the (8} + II daugh- 
ter of No. 10) A is later, although in this group the tumor rate 
is high. 

MEDIXTM BATE TUMOR MICE 

6. While in the large majority of the hybrids between Cream 
and English the tumor rate is intermediate (32.3 per cent) be- 
tween that of the Cream and English strain, in some of them it 
is low like that of the Cream, and in one of them, as we have 
seen, it approaches that of the English strain. We have divided 
these hybrids in two larger groups, one containing the low and 
the other the medium tumor rate strains. We find that the tmnor 
age in both reaches almost, but not quite, that of the English 
parent. Tumors appear relatively early in these groups; only 
in two of these hybrid strains the tumor age is intermediate. 
Again we notice that in small strains comprising only 26 or 27 
individuals the determination of the timior age is no longer 
certain. 

7. The result as to the effect of hybridization on tiunor age is 
different in another hybrid strain in which Cream was crossed 
with the No. 10 (I daughter) strain. While in this case the 
tumor rate is intermediate, the tiunor age approaches that of 
the Cream, although it does not quite reach the lateness of the 
tumor age of the Cream. On the other hand, in a cross between 
Silver and the same No. 10 strain, although the tumor rate is 



ORIGIN OF TUMORS IN MICE 205 

again intermediate— in accordance with the low tnmor rate of 
Silver which almost corresponds to that of the Cream — ^the txmxor 
age is decidedly different and very much resembles that of the 
No. 10 strain; this is in accordance with the fact that, as far as 
can be determined from the restricted number of Silver mice at 
our disposal, the tumor age is in this latter strain very much 
higher than that of the Cream* 

This dissociation between tumor age and tumor rate is like- 
wise apparent in the LoAdon strain, where the tumor age almost 
reaches that of the English, while the txmxor rate is very much 
lower. 

A similar dissociation we find in the following straans: Car- 
ter (German + Carter) B, English (344) + 8i, Cream + (198 
English + (8 + German) F4) and in the Waltzer and English 
as well as the Waltzer and Cream hybrids. In the case of the 
Cream + (198 English + (8 + German) F4) strain the number 
of mice is perhaps too small to permit much consideration. The 
Carter strain is in its origin related to the English; the tumor 
rate in this case stands on the border between a medium and 
high rate. The other strains in this class are hybrids in which 
ei^er English, Carter, or Waltzer mice enter. In these hybrids 
there seems to be a tendency to an early tiunor age. This would 
agree with the early tumor age of the English and Carter parent 
strains. In regard to the pure Waltzers, we have no statistics 
as to their tumor age and tumor rate. To judge from the result 
in the crosses we can only infer that their tumor age was early. 

8. In a mmiber of tumor strains with medium tumor rate 
the timior age corresponds to the rate. (11 to 15 per cent of 
tumor mice in the I age period, an average of 45 to 55 per cent in 
the III age period.) In this group we include: European + 
No. 8, Heitler, Grerman, 8 + German, No. 8, White Cream + 
I daughter of No. 10, Black Cream + European, Vermont Wild 
+ English F« and German + No. 6 F4. Of these. Black Cream 
+ European and Vermont Wild + English stand on the border 
between a medium and late tumor age. 

9. There are some strains with a medium tumor rate and a 
late tumor age. To this group belong: 101 (English) + (Euro- 



206 LEO LOEB 

pean + 103), European + English Tan, 8i + II daughter of 
No. 10, (8J + II daughter of No. 10) + lid. of No. 10, Cream + 
I daughter of No. 10, and European + Cream. We notice that 
the parents (European + 103), European 8|, and Cream have 
the tendency to impart to hybrid strains a late tirnior age. All 
those parent strains have either themselves a late tiunor age 
or their tumor age is on the border between medium and late, 
with exception of the European which, notwithstanding their 
relatively early tumors, likewise seem to impart a late tumor 
age to their hybrid offspring. In this strain there was perhaps 
present a greater individual variabUity m the tendency towards 
a certain- tumor age. We notice furthermore that whenever the 
I or II daughter of No. 10 enters into a cross with a second parent 
strain with a tumor age which is late or approaches lateness, the 
No. 10 component of the cross, despite its high tumor rate, does 
not tend to make the tumor age earlier, because the European 
+ I and II daughter of the No. 10 mice themselves have a rela- 
tively late tumor age considering their high tumor rate. 

10. If we compare the list of the high tumor rate strains with 
that of the medium tumor rate strams, we find in the former the 
tumor age on the whole much more homogeneous than in the 
latter. In the high tumor rate strains the tumor age in all the 
important strains with exception of strains into which the daugh- 
ters of No. 10 enter is early, while in the medium tumor rate 
strains the variability as to tumor age is much greater, although 
the deviations are about equally distributed on both sides of a 
medium tumor age, which latter represents the average. 

LOW STRAIN TUMOR MICE 

11. The Cream strain is typical of the low rate tumor strains. 
The difiference in the tumor age between this stram and a typi- 
cal high rate strain is striking; the tumors appear here much 
later in life than in the high tumor rate strain. We must of 
course take into consideration the fact that the smaller the 
number of tumor mice, the greater is the effect on the percentage 
age distribution of a small variation in the number of mice dying 
in the different age periods. Therefore we should expect a 



ORIOm OF TUMOBS IN MICE 



207 



Low rate tumor mice 



STBAIH 



Total English Cream low rate 
hybrids 

(344 + Black Cream) Fi + 
Cream (white) 

(English 344 + Cream, new 
individuals). . . ^ 

(English Sable 4444 + Cream) 

Total Cream: 

(Total Cream X) 

(Cream X A) 

(Cream X B) 

(Cream X C) 

(Old Cream) 

(Cream A, mostly black) 

(Cream B, mostly white color) 

(Among Cream B, June 1014 
Family) 

(The rest of Cream B after 
deduction of June 1914 fam- 
ily) 

(Cream Black II) 

(Cream Black III) 

(Cream Y) 

(Total new Cream) 

Total European 

(European A) 

(European B) 

Total European + 102 or 103. . 

(European + 102 or 103 A) 

(European + 102 or 103 B). . . . 

(London 481 ) 

Total London + (European + 
103) F, 

(London + (European + 103) 
F,A] 

[London + (European + 103) 
F.B] 

415 » [101 English + (Euro- 
pean + 103)1 

Total No. 8J 

(No.SJA) 

(No.SJB) 



svu- 




FEDCEMTAGB Or TUIIOBS Ilf 


BEROr 


TUMOB 




MICE 


BATB 


I age period 


II age period 


III ace period 




per cent 


per cent 


per cent 


per cent 


318 


5.7 


28.0 (2.5) 


38.0 (3.4) 


34.0 (3.0) 


11 


0.0 


0.0 


0.0 


0.0 


143 


9.0 


23.4 


37.0 


39.6 


175 


2.8 


30.1) 


36.0 


34.0 


878 


5.9 


8.0 (1.0) 


26.0 (3.3) 


66.0 (8.2) 


260 


2.7 


18.2 (0.8) 


52.3 (2.3) 


29.5 (1.3) 


135 


4.0 


27.0 (1.4) 


73.0 (3.8) 


0.0 (0.0) 


77 


2.6 


0.0 (0.0) 


33.0 (2.0) 


67.0 (4.0) 


48 


0.0 


0.0 


0.0 


0.0 


226 


2.0 


0.0 (0.0) 


20.0 (1.2) 


80.0 (4.3) 


114 


8.0 


10.4 (1.7) 


23.0 (3.7) 


67.0(11.0) 


80 


19.0 


3.0 (1.1) 


39.0(14.0) 


58.0(21.0) 


40 


35.0 


4.6 (2.5) 


33.5(18.0) 


61.9(33.0) 


49 


6.0 








77 


4.0 


13.0 (1.3) 


0.0 (0.0) 


87.0 (8.5) 


98 


11.0 


13.6 (3.0) 


14.0 (3.1) 


72.4(16.0) 


14 


0.0 


0.0 


0.0 


0.0 


349 


10.5 


7.6 (1.7) 


25.0 (5.6) 


67.4(15.0) 


263 


9.0 


21.0 (3.4) 


34.0 (5.4) 


45.0 (7.4) 


113 


16.0 


21.0 (6.0) 


42.0(12.0) 


37.0(11.0) 


150 


3.3 


18.0 (1.3) 


14.0 (1.0) 


68.0 (5.0) 


221 


15.0 


3.0 (1.0) 


24.0 (7.5) 


73.0(23.0) 


146 


20.5 


4.3 (1.5) 


23.0 (8.0) 


71.7(25.0) 


75 


5.3 


0.0 (0.0) 


30.5 (5.5) 


69.5(12.5) 


43 


0.0 


0.0 


0.0 


0.0 


148 


7.0 


6.0 


25.0 


69.0 


85 


5.0 


13.0 (1.2) 


30.0 (2.8) 


57.0 (5.4) 


63 


8.0 


0.0 (0.0) 


31.0 (4.1) 


69.0 (9.0) 


80 


12.5 


12.0 (3.5) 


41.0(12.0) 


47.0(3.5) 


241 


13.0 


11.0 (2.9) 


50.0(16.0) 


39.0(10.0) 


158 


17.0 


13.3 (4.0) 


53.0(16.0) 


33.7(10.0) 


28 


0.0 


0.0 


0.0 


0.0 



208 



LEO LOEB 



WTRAUS 



(No.SiC) 

Total (European + 102) Fi -f 

8JF4 

[(European + 102) Fi + 8i 

F«A] 

[(European + 102) Fi + 8 J F4 

B] 

Total (European -f 108) Fi + 

Illdof No. 10 

[(European + 103) Fi + Hid 

of No. 10 A] 

[(European + 103) Ft + Hid 

of No. 10 BJ 

[(European + 103) Ft + Hid 

of No. IOC] 

English SOver 

Engliflh Silver Fawn 

London + Silver 

(N0.8B) 

(8J + lid of No. 10 B) 

German + Carter A 

English Sable 344 + (European 

+ English 146) (344-146). . . . 
White Cream + (European 

161 + Id of No. 10 = No. 607) 
Cream + (Cream + European 

428)Fi 

German + No. 8 



IRTM- 

BBBOr 

MZCS 



55 

602 
473 
129 
219 
168 
30 

21 
127 

67 
176 

30 

28 
358 

21 

128 

11 
112 



TUMOB 
BATB 



per etnt 

9.0 

16.0 

16.0 

15.0 

14.6 

17.0 

13.0 

0.0 
7.0 
12.0 
12.0 
10.0 
14.0 
9.0 

5.0 

9.5 

10.0 
0.0 



PBBCBMTAOB OF TUMOBB Of 



lageiMriod 



per etmt 
7.6 (2.0) 

6.5 

7.0 

8.6 

4.4 

6.3 

0.0 

0.0 
135.0 (3.5) 



(0.0) 



0.0 
8.0 
0.0 
0.0 
6.6 

0.0 



5.8(1.65) 

0.0 
0.0 



11 ■<» period 



pmr offtt 
33.0 (9.0) 

21.0 

18.0 

28.5 

29.0 

28.0 

10.7 

0.0 

65.0 (6.6) 
38.0(11.0) 
24.0 

0.0 

100 
26.0 

100 

17.5 (5.0) 

100 
0.0 



macs period 



P0r offil 
59.5(16.0) 

72.5 

75.0 

62.9 

66.6 

65.7 

89.3 

0.0 

0.0 (0.0) 
62.0(18.0) 
68.0 

100 

0.0 
67.4 

0.0 

76.7(22.0) 

0.0 
0.0 



much greater variability in the age distribution in the low rate 
than in high rate tumor strains. This does not, however, apply 
in the case of the representative strains of this group because we 
are working with a very great number of individuals in these 
strains. Thus, while in the Cream strain the tumor rate is as 
low as 5.9 per cent, the strain consists of 878 mice and among 
them are 52 tumor mice, a sufficiently large number for our 
purpose. If we consider, on the other hand, the substrams in 
this group, then the number of tumor mice usually becomes so 
small that the development of one or two more tumor mice in 



OBIOm OF TUMORS IN MICB 209 

an earlier age period may change considerably the age distribu- 
tion. Notwithstanding this fact we find in the Cream strain 
that the smaller substrains behave essentially like the main 
strain; we notice a low percentage of tumor mice in the first age 
period and a high percentage in the third age period, with ex- 
ception of the Cream X A substrain where we have to deal with 
only 5 tumor mice, two of which died in the first and three in the 
third age period, and even here a change in time of death in the 
case of one or two individuals would have made the age curve 
typical. 

12. What applies in the case of the Cream strain applies also 
to the other more important low rate tumor strains, as for in- 
stance the European + 102 or 103 where, as a result of the some- 
what higher tumor rate, we have to deal with 34 tiunor mice. It 
also applies in the case of London + (European + 103) Ft, in 
the large strain (European + 102) Fi + 8^ F4, where we can 
base our calculations on 97 tumor mice; furthermore, in the case 
of the (European + 103) + III daughter of No. 10, London 
+ Silver, German + Carter A in which the tumor rate was only 
9 per cent and the age distribution was accordingly typical for 
the low rate tumor strain, while in the German + Carter B, where 
the tumor rate rose, the percentage distribution in the different 
age periods also changed in the typical manner. 

The same age distribution we find furthermore in the White 
Cream + (European 151 + I daughter of No. 10) and some 
smaller strains. In the German + 8 tumors did not develop. 
Wherever the European + 102 or 103 enter a cross, they not 
only depress the timior rate, but call forth the typical age dis- 
tribution characteristic of the low tumor rate strains. This 
applies even in the case of the English 101 + (European + 103) 
where the association with the English strain succeeded in rais- 
ing the tumor rate to 34 per cent and thus in producing a medium 
rate tumor strain; but the tumor age is very nearly as late as 
that of the Cream. 

13. There are a few groups in this class in which the timiors 
appear earlier, namely, European, a small substrain 415 » 101 
English + (European + 103), No. 8i, and English silver. Eng- 



210 LEO LOEB 

lish Silver is related to the English and the tumor age ap- 
proaches that of the English. Notwithstanding the relatively 
small number of tumor mice in this strain, this age distribution 
is probably not altogether due to a coincidence; at least we may 
interpret in this sense the fact that even in the hybrids in which 
English Silver enters the tumor age is relatively early. In 
No. 8J and in 415 = 101 + (European + 103) the timior age 
stands at the border between the early and medium types, or 
rather it approaches the mediimi type. In the European strain 
the tumors appear earUer. Here we may have to deal with 
another mstance in which tumor rate and tumor age are dis- 
sociated; in the case of the low rate tumor mice the ntmiber of 
tumor mice is, however, relatively so small that in individual 
strains, except the largest ones, chance variations cannot be 
excluded with certainty. 

14. In the English-Cream hybrid with a low tumor rate the 
tmnor age is very early, similar to that of the English. The 
number of mice is 318, the tumor rate is 5.7 per cent; we have 
to deal with 18 tumor mice. This corresponds to the similar 
high tumor age in the English-Cream hybrids with medium 
tumor rate. In this group the number of mice is still greater 
and our classification is based on the cancer age of 239 tumor 
mice. 

We may therefore conclude that in the English-Cream hybrids, 
independently of the ttmior rate which may vary in different cases, 
the tumor age approaches that of the English. This applies 
even in the case of the two individual English-Cream hybrids 
with a low tumor rate. 

In the case of the English-Cream hybrids we find, therefore, 
a splitting in the inheritance of tumor rate and tmnor age. 

THE TUMOR AQE OF THE HIOH, AND LOW TUMOR RATE MEDIUM 

STRAINS 

15. The great difference in the tumor age of the high, medium, 
and low tumor rate strains is brought out very clearly if we dis- 
tribute aU our strains among three great groups, according as to 



ORIGIN OF TUMORS IN MICE 211 

whether their tumor rate is high, medium, or low, treating each 
group as a whole and determining in each of them the tumor age 
of the group. 
Thus we obtain the following figures: 

Total of the high tumor rate strains 

I age period. 2741 mice 885 tumors » 32.3% 22.8% 

II age period. 1176 mice 636 tumors - 54.1% 37.3% 

III age period. 225 mice 132 tumors « 58.7% 40.4% 

Average timior rate 60.3% 

Total of the medium tumor rate strains 

I age period. 4351 mice 394 tumors » 9.0% 14.6% 

II age period. 2855 mice 635 tumors - 22.2% 36.0% 

III age period. 1238 mice 374 tumors - 30.5% 49.4% 

Average tumor rate 32.2% 

Tot€d of the low tumor rate strains 

I age period. 3971 mice 67 tumors - 1.60% 8.5% 

II age period. 2747 mice 152 tumors - 5.53% 27.8% 

III age period. 1362 mice 173 tumors - 12.7% 63.7% 

Average tumor rate 9.9% 

This arrangement brings out very clearly the difference in the 
tumor age of the different groups. In the group composed of 
the high tumor rate strains the average tumor rate is 60.3 per cent, 
the tumors appear early, almost, but not quite, as early as in 
the English strain. The figures for the I, II, and III age period 
are I 22.3 per cent, II 37.3 per cent. III 40.4 per cent. In the 
second group the average tumor rate is 32.3 per cent. The 
tumor age is correspondingly intermediate. The figures for the 
3 age periods are: I 14.6 per cent, II 36 per cent. III 49.4 per 
cent. In the third group the average tumor rate is 9.9 per 
cent. The average tumor age is approximately that of the 
Cream; the figures are I 8.5 per cent, II 27.8 per cent. III 63.7 
per cent. We may therefore conclude that the higher the tumor 
rate in a strain, the earlier the tumors appear on the average; 
the greater is the number of tumors in the first age period, and 
the smaUer is the number in the third age period. The differ- 
ences in the first age period are, however, somewhat greater 



212 LEO LOEB 

than those m the third age period. The differences between the 
figures for the II age period, which represent the turning point 
are, as might be expected, much smaller in the three groups. 

16. This relationship between tumor rate and tumor age can 
be shown in still another way. We may arrange all the strains 
in three different classes in accordance with their tumor age, 
the first class comprising those strains having a tiunor age simi- 
lar to that of the high tumor rate group, the second comprising 
those strains having a tumor age similar to that of the medium 
tumor rate group, and the third comprising the strains with a 
tumor age similar to that of the low tumor rate group. 

We then find in the first class, comprising the strains with an 
early tumor age, 19 strains with a high tumor rate, 10 strains 
with a mediimi timior rate, and 3 strains with a low timior 
rate. Almost 60 per cent of the strains belong, therefore, to the 
high tumor rate strains. 

In the second class, with a medium timior age, we find 10 
strains with a medium tumor rate, 2 strains with a low tiunor 
rate, and 5 strains with a high timior rate. In this class 59 per 
cent of the strains belong to the medium tumor rate strains. 

In the third class, with a late timior age, we find 14 strams 
with a low tumor rate, 6 strains with a medium tumor rate, and 
1 strain with a high tumor rate. Sixty-seven per cent of the 
strains belong to the low tumor rate strains. In each class the 
majority of strains have a tumor rate which corresponds to the 
tumor age of that class. 

ON THE PERIOD OF UFB IN WHICH THE DEATH-BATE FROM CANCER 

IS HIGHEST IN MICE 

17. It is generally assumed that in man the cancer incidence 
on the whole increases with increasing age. However, we find 
certain variations according to the kind of cancer which is con- 
sidered ; thus the death-rate from mammary and uterine cancer 
reaches a somewhat earlier maximum than the death-rate from 
cancer in general. The same applies to sarcoma. In mice 
Murray finds the death-rate from mammary cancer to reach a 
maximum at the age of thirteen to sixteen months. Murray's 



ORIGIN OF TUMORS IN MICE 213 

statistics differ from our own in that they are based on a much 
smaller number of mice and in the method of computation he 
uses. Later we shall apply the same method also to our material* 
We found that the age of death from mammary cancer in 
mice is not a fixed point, but varies with the cancer rate. The 
cancer rate being variable in different strams, the cancer age 
is likewise variable. In general we may state that the cancer 
age is the earlier, the higher the cancer incidence. 

18. If we consider the total figures for high, medium, and low 
timior rate strains respectively, we find that in the high tumor 
rate strains the difference between the cancer rates in the third 
and second age periods is very small, but the cancer rate is slightly 
higher in the third age period. In mice above the age of eight- 
een months the cancer rate is therefore slightly higjher than in 
mice between the age of twelve and eighteen months. The 
maximum still is in the third age period. 

In the medium tumor rate strains the difference between the 
cancer rate in the second and third age periods is more marked; 
the maximum is here quite definitely in the third age period. 
In the low tumor rate strains the difference between the tumor 
rate in the second and third age periods is still much greater; 
the maximum is very decidedly in the third age period. The 
lower the tumor rate, the more decidedly the maximum moves 
to the third age period. The lower the tumor incidence, the 
greater the maximum reached in the third age period and the 
greater the difference between the tumor incidence in the second 
and third age period. The difference between the tumor rates 
in the first and second age periods varies less; it happens to be 
least in the group of the high rate tumor strains. 

19. If instead of considering the groups as a whole, we con- 
sider the individual strains composing them, the difference in 
the situation of the maximum .tumor incidence in different strains 
comes out still more clearly. Thus in the 8i + 328, a high tumor 
rate strain, the maximum is decidedly in the second age period, 
while in the English, although they have a slightly higher tumor 
rate, the maximum is somewhat higher than in the 8^ + 328, 
although it is still in the second age period if we consider the 

TBB 30VXSAL OF CAMCSB BUXABCH, VOL. VI, NO. 3 



214 LEO LOEB 

total strain; the difference between the second and third age 
periods is however very small in this case and in the substrains 
we find some variation as to the maximum. 

The great majority of the English hybrid strains behave 
like the English strain as far as the maximum of the tumor in- 
cidence is concerned. Similarly to the English strain behave in 
this respect particularly the English-Cream hybrids and Euro- 
pean + EngUsh TaDi, irrespective of their tumor rate. 

In addition to various English hybrids, the strains London, 
8 + German, and 8| show a maximum similar to that of the 
English mice. On the other hand, in the European 151 + I 
or II daughter of No. 10, in which the tumor incidence is very 
high, the maximum is decidedly in the third age period. 

We must then conclude that the maximum of tumor incidence 
is found at a different age level in different strains, and that this 
peculiarity is transmitted by heredity in the same way as the 
other characteristics of the strain. 

The data as to the age period in which the maximmn cancer 
incidence is found in mice, which we present here, permit only 
comparative conclusions. We are at present operating with 
three age periods which are not equal in length and we cannot 
therefore state at which age period the maximum occurs. But 
•our data have a relative value. They show the shifting of the 
age period in different strains in dependence upon the variations 
in the cancer rate of the various strains and groups. And we 
may furthermore conclude that in all statistics so far no account 
.has. been taken of this variability of the tumor age iu general 
and of the maximum in the tumor rate in particular. All avail- 
able statistics give an average, in which this variable factor is 
disregarded and in which necessarily the result will differ in 
accordance with the relative preponderance of high or low tumor 
rate families or strains. 

20. If instead of considering the three groups of mice sepa- 
rately we determine the tumor iacidence in the three age periods 
in all the mice, irrespective of the group to which they belong, 
we obtain the following figures: In the first age period there 
were 11,063 mice; 1346 mice developed tumors during this period 
a 12.2 per cent. 



OBIOIN OF TUMOBS IN MICE 215 

In the second age period there were 6778 mice with 1423 
tumor mice "21 per cent. 

In the third age period there were altogether 2825 mice, with 
679 tmnor mice » 24 per cent. 

If we calculate these figures on the basis of a tumor incidence 
of 100 per cent, we find for the first age period a tumor incidence 
of 21 per cent, for the second age period an incidence of 36.7 per 
cent, and for the third age period an incidence of 42.3 per cent. 
These figures are somewhat between those of the high and medium 
tumor rate mice, but they approach very closely those of the 
high tumor rate mice. The maximum of the tumor incidence 
is accordingly in the third age class, but the difference between 
the figures for the second and third age period is relatively small. 

21. If instead of considering in each age period the number 
of mice alive at the begiiming of this period and the number of 
mice developing timiors in this period, and on this basis de- 
termining the percentage figures of tumor incidence for each 
age period, we determine what percentage of mice will develop 
cancer during the whole period following this date irrespective 
of the number of mice eliminated through death during each 
period, we obtain a different result. We had altogether 11063 
mice which reached the age above six months. Of these 3448 
subsequently developed cancer = 31.2 per cent; of 6778 mice 
which reached the age of twelve months, 2102 later developed 
cancer » 31 per cent; and of 2825 mice reaching the age of 
eighteen months, 679 developed cancer » 24 per cent. Figured 
out on this basis the tumor rate is similar for mice reaching the 
age of six and of twelve months. In these groups the tumor 
rate is higher than in mice reaching the age of eighteen months. 

This mode of computation seems to us less satisfactory than 
the first metiiod used, because it includes the number of indi- 
viduals dying in each period among those which are liable to be 
affected by cancer in subsequent periods, and it thus makes the 
tumor incidence in the first two age periods too high. 

22. If we leave out of consideration the mice dying below the 
age of 6 months we find among our material altogether a tumor 
incidence of 31.2 per cent. This is a relatively high tumor rate 



216 LEO LOEB 

which presumably exceeds the average tumor rate among white 
mice in imselected strains. Our material is a selected one, in 
which a special value was attached to high tumor rate strains. 

23. If we determine with the second method the cancer inci- 
dence in the various groups of high, mediiun, and low tumor rate 
mice, we find the following figures. 

High tumor rate mice 

I 2741 mice 1653 tumor mice » 60.3% 

II 1176 mice 768 tmnor mice » 65.3% 

III 225 mice 132 tumor mice » 58.7% 

Medium tumor rate mice 

I 4851 mice 1403 tumor mice -^ 32.3% 

II 2855 mice 1009 tumor mice -^ 35.8% 

III 1238 mice 374 tumor mice - 30.2% 

Low tumor rate mice 

I 3971 mice 382 tumor mice >■ 9.9% 

n 2747 mice 326 tumor mice -^ 11 .7% 

III 1802 mice 173 tumor mice « 12.7% 

On this basis the percentage figures for the different age periods 
are very similar in the three groups. But in the high and 
medium tumor rate mice the maximum of the tumor incidence is 
in the second age period, in the low tumor rate mice it is again 
in the third age period. Even this method, which is more 
faulty than the first one, brings out the shifting in the tumor 
incidence with varjdng tumor rates. 

THE INHEBTTANCE OF TUMOR AGE AND INCIDENCE IN HYBRIDS 

24. We have already referred to the behavior of individual 
hybrid strains as far as the relation between tumor age and tumor 
rate is concerned. We have also referred to some of the larger 
groups of hybrids. We shall now discuss the behavior of hybrids 
in some of the lai^er groups connectedly and compare their 
tumor rate and age with that of the parent strains. 

(a). EngliahrCream hybrids. The figures for the parents are 
as follows: 



ORIGIN OF TUMORS IN MICE 217 

A. Total English 

I 689 mice 280 tumor mice - 40.6% 25% 
II 258 mice 161 tumor mice - 62.5% 39% 67.6% 
III 46 mice 25 tumor mice - 54.5% 36% 

B. Total Cream 

I 878 mice 9 tumor mice » 1.0% 8% 

II 600 mice 20 tumor mice - 3.3% 26% 5.9% 
III 279 mice 23 tumor mice « 8.2% 66% 

C. Total English-^ream Hybrids 

I 1057 mice 92 tumor mice » 8.6% 20% 

II 668 mice 120 tumor mice » 18.0% 42% 24.3% 
III 280 mice 45 tumor mice » 16.1% 38% 

In these hybrids the tumor rate is on the whole mtennediate 
between those of the parents, although somewhat nearer the Cream 
parent than the English. The tumor age, on the other hand, is 
almost that of the English parent. In this case we have to deal 
with such large figures that we are justified in assuming that 
these figures express causal relations and are not due to accidental 
findings. 

The same relation comes out still more clearly if we divide the 
English-Cream hybrids into three subgroup^. 

A. Total English-Cream Hybrids with a medium tumor rate 
(and with a slight admixture of high rate tumor mice) 

I 434 mice 53 tumor mice » 12.0% 20% 

II 264 mice 67 tumor mice » 25.0% 41% 32% 

III 100 mice 21 tumor mice » 23.0% 39% 

B, ToUU English'4Jream Hybrids with a low tumor rate 

I 318 mice 8 tumor mice « 2.5% 28% 
II 205 mice 7 tumor mice = 3.4% 38% 5.7% 
III 101 mice 3 tumor mice » 3.0% 34% 

In this case the tumor rate is altogether that of the Cream, 
but again the tumor age is that of the English parent strain. 

C. Total English Tan (121) + Cream 

I 267 mice 32 tumor mice » 12% 22% 
II 173 mice 36 tumor mice » 21% 38% 31% 
III 63 mice 14 tumor mice » 22% 40% 



218 loco LOSB 

Also in this strain of hybrids the tumor rate is intermediate 
while the tumor age is that of the English parent strain. 

b. The same conclusion, that factors for tumor rate and age 
may be transmitted independently of one another, follows from 
a consideration of the hybrids between Cream and European 
151 + I daughter of No. 10. 

The tumor rate and tumor age of the Cream we have already 
given. 

Pigure9 for the moiher strain European 161 + I daughter of No, 10 are as foUowe: 

I 254 mice 03 tumor mice » 25.0% 15.0% 
II 148 mice 120 tumor mice •> 81.0% 88.0% 72.0% 
III 46 mice 30 tumor mice - 88.0% 52.0% 

The Total Hybrids Cream + (European + 1 daughter of No. 10) 

{three different croeeee used) 

I 408 mice 22 tumor mice » 4.4% 8.4% 
II 365 mice 53 tumor mice - 15.0% 29.0% 27.5% 
III 188 mice 62 tumor mice « 33.0% 62.6% 

In this case the tumor rate is again on the whole intermediate 
and sEghtly nearer that of the Cream, but the tumor age is that 
of the Cream, contrary to what we foxmd in the former hybrids. 
If we omit from the table of hybrids one cross in which the tumor 
rate was low, the figures for the Cream + 10 hybrids with medium 
tumor rate (2 crosses) are as follows: 

Total Cream + (European + 1 daughter of No, 10) hybrids with medium tumor rate 

I 370 mice 20 tumor mice « 5.4% 10% 
II 286 mice 40 tumor mice - 17.0% 32% 34% 
III 161 mice 56 tumor mice - 31.0% 58% 

Again the tumor rate is intermediate and the tumor age is 
almost that of the Cream. In this connection it is of interest 
to remember that the tumor age of the European 151 + I or II 
daughter of No. 10 is later than that of the English, and accord- 
ingly it has a tendency to transmit to crosses a later tumor age 
than the English. 

If we cross strains in which tumor rate and tumor age differ 
less than in the preceding strains, we again obtain a tumor rate 



ORIQIN OF TUMOBS IN MICE 219 

and tumor age intermediate on the whole, or perhaps resembling 
slightly more one parent than the other. This is shown in the 
following crosses: 

A, Oerman 

I 42 mice 8 tumor mice « 19% 15.5% 

II 10 mice 7 tumor mice » 37% 30.3% 40.5% 
III 3 mice 2 tumor mice -> 66% 54.2% 

In this case the number of mice is rather small; we can there- 
fore use it only with caution. 

B. Total No. 8 

I 243 mice 17 tumor mice -> 7% 12.0% 
II 150 mice 24 tumor mice *- 16% 27.4% 27.5% 
III 78 mice 26 tumor mice - 35% 60.6% 

C. Total No. 8 + Chrman 

I 373 mice 30 tumor mice "-8% 14.0% 
II 280 mice 66 tumor mice - 24% 42.0% 34.0% 
III 123 mice 31 tumor mice - 25% 44.0% 

The tumor rate of the hybrids is about intermediate and the 
tumor age is very similar in both parent strains as well as in the 
hybrids. 

d. Another strain may be of interest; this resembles in its 
character the last named hybrid strain and is related to it. We 
crossed the 8 + German with an English male, the offspring of 
Tumor mouse 198. 

The data for the parent strains have already been given. The 
figures for the cross are as follows : 

BnglUh (198) + (8 + Oermon) 

I 217 mice 61 tumor mice - 28% 20% 

II 89 mice 52 tumor mice -> 58% 41% 54% 

III 9 mice 5 ttunor mice -> 55% 39% 

In this case the tumor rate of the hybrids is again intermediate, 
but slightly nearer that of the English. The tumor age is also 
intermediate, but again somewhat nearer that of the English. 
In this case no distinct split between tumor rate and tumor age 
occurred in the hybrids. 



220 LEO LOEB 

25. We believe that in these investigations we have established 
a definite relation between tumor incidence and tumor age. • In 
general the average tumor age is the earliest in those strains of 
mice in which the niunber of individuals affected by cancer of the 
breast is greatest. There is in addition a diflference in the tumor 
age in a niunber of strains in which the tumor rate is similar; 
or, conversely, in a niunber of strains in which the tumor age is 
similar the tumor rate may differ. We have seen furthermore 
that in crosses tumor age and tumor rate may be inherited in- 
dependently of each other, or in other cases the tmnor age may cor- 
respond to the tumor rate. How can these facts be interpreted 
in accordance with the current theories of heredity? We may 
assume as most probable that the tendency to cancer depends on 
the presence of multiple factors. 

A certain niunber of these factors must be present in an indi- 
vidual if a tumor is to appear. These factors determine the in- 
tensity of the tendency towards cancer in each individual; but this 
intensity tends not only to cause a tumor to appear in a certain 
animal, but it has the tendency to make it appear at an early 
period of life. Tumors appear the earlier, therefore, the greater 
the number of tumor factors present or the greater the impor- 
tance of the factors represented. We may assume that different 
strains differ greatly in the number and character of these fac- 
tors present in the average of the individuals. In the English 
and 8^ + 328 there may be in an individual affected with can- 
cer, on the average, a greater number of the more effective fac- 
tors than in a Cream individual which happens to have just a 
suflBcient minimum of factors to insure the appearance of a tumor. 
Therefore the intensity in the tendency towards the appearance 
of tumors is greater in the English and 8J + 328 than in the 
Cream. In all these individuals tumors appear, but the greater 
intensity in the English and 8^ + 328 causes the tumors to 
appear earUer in life in these latter strains. There are, of course, 
variations in the individuals of different strains, and some indi- 
viduals of the 8| + 328 or English may resemble the average 
Cream as far as the character and number of their timior fac- 
tors are concerned. There may be, in addition, among the 



ORIGIN OF TUMORS IN MICE 221 

multiple factors determining tumor growth, some special factors 
which determine more directly the tumor age, and these special 
factors may be inherited independently of the multiple factors 
which determine mainly the appearance of a tmnor. 

These special factors may overcome the effect of the ordinary 
tumor factors-, which in themselves already have a tendency to 
affect the tiunor age. If, on the other hand, we assume that 
factors determining tumor age are always distinct from those 
determining tumor incidence, then we must conclude that both 
sets of factors are usually linked to each other in such a way that, 
in general, in those strains in which the tumor rate is high the 
factors determining an early appearance of tumors are also pres- 
ent; but that it is possible to dissociate these two sets of factors 
in certain cases, especially through hybridizations. Both in- 
terpretations are related to each other; the first interpretation 
including the second one. These interpretations must at present, 
of course, be considered as merely of atentative character, but 
they seem best to represent the facts so far established. 

CONCLUSIONS 

1. The tumor age of a certain strain is as definitely determined 
by heredity as the tumor incidence; the tumor age may be a finer 
means of distinction between different strains than the tumor 
incidence. 

2. There is a definite relation between timior rate and tumor 
age in mice. In those groups or strams in which the tumor 
incidence is great the tumors tend to appear early, and in those 
groups or strains m which the tumor rate is low the tumors 
tend to appear late. We can arrange our mice in three groups, 
with high, mediiun, and low tumor rate; in these three groups 
the tiunor age shows corresponding changes. 

3. The period of life at which a certain kind of tumor shows a 
maximiun frequency in a certain species is not definitely fixed, 
but it varies with the rate of tumors in certain strains. The 
usual statistics represent an average between the maxima in 
different strains in which the maximum varies in accordance 
with their tumor rate. 



222 LEO LOEB 

4. In addition to this general relation between tumor age and 
tumor rate^ there is in certaiQ strams a specific tmnor age which 
may differ from that expected in the strain on the basis of its 
tumor rate. Furthermore, in hybrid strains tumor rate and 
tumor age may be inherited independently of each other. 

5. These relations between tumor age and tumor rate can 
best be explained if we assume that the hereditarily transmitted 
constitution, so far as it represents the tendency of the organism 
to develop tmnors, depends on the codperation of multiple fac- 
tors. These multiple factors determine the intensity in the 
tendency to tumor development in a certain individual. In 
general, the greater is this intensity, the earlier do the timiors 
appear and the greater is the probability that in related indi- 
viduals there exists likewise a tendency to the development 
of tumors. 

It is furthermore probable that in addition to the general 
factors determining the intensity in the tendency towards the 
development of cancer, there exist factors which determine 
specificaUy the tumor age in certain individuals and strains. 

REFERENCES 

(1) Lathbop, a. £. C.| AND LosB, Lbo: Proc. Soo. Ezper. Biol, and Med., 1913, 
xi, 34; Jour. Exper. M., 1915, zxii, 646 and 713, 1918, xxviii, 475; 
Jour. Cancer Res., 1919, iv, 137. 



FURTHER INVESTIGATIONS OF DISTURBANCES OF 
BLOOD SUGAR EQUILIBRIUM IN THEIR 

RELATION TO NEOPLASIA* 

G. L. ROHDENBURG, O. F. KREHBIEL, and A. BERNHARD 

From Columbia University, InatittUe of Cancer Research, founded by George Crocker , 
F, C, Woodf Director, and the Letboralory of the Lenox Hill Hospital 

Received for publication January 16, 1922 

In previous publications (1), it has been shown that the 
injection of a protein into the animal body is followed by dis- 
turbance of the blood sugar eqiulibriunii and that this disturb- 
ance of sugar equilibrium can be influenced (either increased 
or decrease) by the ablation of one or more of the endocrines 
(2). The present report concerns a further study of: (a) the 
interrelationship of the endocrines in the regulation of blood 
sugar equilibrium; (6) the relation of the degree and type of 
blood sugar disturbance after the injection of protem to the 
strength of the anti-bodies developed; (c) the relation of the 
type and intensity of the reaction following the injection of 
homologous protein to neoplasia in man. 

I. EXPERIMENTAL PART 

The technical steps of the method employed are simple. 
The animal to be tested is starved for a period of twelve hours. 
Just before, and again sixty minutes after the subcutaneous 
injection of the antigen chosen, the blood sugar values are 
determined in milligrams per 100 cc. The lower of the two 
values is subtracted from the higher, the difference being taken 
to indicate the degree of disturbance as measured in milligrams. 
In the present work, in determining the blood sugar values 

* We beg to express our appreciation to all of the attending physicians and 
surgeons of the Lenox Hill Hospital for the clinical material contained in this 
report. 

223 



224 R0HDENBUR6, KREHBIEL AND BERNHABD 

we have used either the method of Epstein or of Kliner, or the 
Wallace and Gallagher modification of the Folin-Wu method. 
We have modified the Wallace and Gallagher method in that 
the blood is measured instead of weighed, 0.1 cc. being used. 
For this purpose a standardized pipet, so made that the capillary 
bore necessary to contain the specified amount is about 15 cm. 
long, was employed. It is essential that the same worker 
make aU determinations m a given experiment and that the 
individual be skilled in the technical steps of the method chosen. 
The same method for determining the sugar values should be 
used throughout any one experiment. These conditions have 
been complied with in the work recorded here. 

In our early experiments it was found that in normal animals 
after the injection of a given antigen, different types and degrees 
of reaction occurred. For example, one animal might show an 
increase of 100 mgm. in the blood sugar, while a second treated 
exactly as was the first might show an increase of but 15 mgm. 
A third animal, on the other hand, instead of showing an increase 
might show a decrease, for example, of 25 mgm. Hereafter, 
we shall use the terms ''plus type'' and "minus t3rpe" to desig- 
nate, respectively, reactions in which the sugar concentration 
rises, and reactions in which a decrease occurs. 

In investigating these variations the first factor considered 
was the influence of sex and age upon the tjrpe of reaction. 
A number of yoimg female, yoimg male, old female, and old 
male rats were tested in accordance with the method previously 
outlined, the injected antigen being 0.5 cc. of a 1 per cent heterol- 
ogous protein (beef peptone). The averaged results of this 
experiment show that in the old males the plus and minus reactions 
occurred with the same degree of frequency, while in young 
and old females and in yoimg males the plus reactions slightly 
predominated. 

In a second experiment, one or another of the endocrine 
glands was removed, the test being performed at varymg inter- 
vals after ablation. A 1 per cent heterologous protem (beef 
peptone), in doses of 0.5 cc, was first used as antigen, and after 
an interval of several days a similar dose of a similar strength 



BLOOD SUGAR EQUILIBRIUM AND NEOPLASIA 



225 



of homologous protein (rat spleen) was employed. The glands 
removed were as follows: both testes complete, both ovaries 
complete, spleen complete, thymus complete, thyroid and em- 
bedded parathyroids about 95 per cent entire; both adrenals 
complete; pancreas about 90 per cent entire. As controls, 
normal animals and animals from which one kidney had been 
removed were used. There were from 6 to 12 animals in each 
group. 

The data of the experiments are given in table 1. The per- 
centages of plus reactions are about the same in the normal 
controls, in those with one kidney removed, and in the thymus- 

TABLE 1 



OBOAN BBIfOVXD 



Normal controls 

Kidney 

Thyroid 

Thymus 

Spleen 

Adrenals 

Pancreas 

Testes 

Ovaries 



PBBCXNTAOa oxviKa vLvn 
. BBACnONS 



Homolopoua 
protein 



70 
60 



39 

83 

100 

58 


78 



Heterolocouft 
protein 



65 
55 

100 
20 
16 

100 

67 



75 



PXRCBMTAOB OnrXNO MIKUB 
BBACnOMB 



HomolofEous 
protein 



30 
40 

100 

66 

16 



42 

100 
22 



Heterokcous 



to^loBoi 
protein 



35 
45 


80 
83 



33 

100 

25 



free, pancreas-free, and ovary-free groups, irrespective of which 
antigen was used. There was a difference in the types observed 
in the spleen-free group, for of these 16 per cent gave a plus 
reaction when injected with heterologous protein, as compared 
with 83 per cent giving a plus reaction when homologous pro- 
tein was used as antigen. One type of reaction only was ob- 
served in three groups: the adrenal-free and thyroid-free groups 
gave 100 per cent plus reactions, and the testes-free 100 per 
cent minus reactions when injected with heterologous protein. 
When homologous protein was used as antigen the adrenal-free 
and testes-free groups reacted in the same manner, while the 
thyroid-free groups gave 100 per cent minus instead of plus 
reactions. 



226 



BOHDENBUBO, KBBHBIEL AND BBBNHABD 



EEaving observed the reactions after single gland removal, 
and in that manner having obtained groups in which all the 
animals reacted aUke as regards type, we studied, in a third 
experiment, the effect of temporary overactivity of various 
glands. Extracts^ in the proportion of 5 grams of the dried 
powdered gland in 25 cc. of phjrsiological saline solution, were 
used as antigen, the dose being 0.6 cc. In this fashion extracts 
of the pituitary, pancreas, adrenal, and thyroid glands of sheep 
were prepared, and. because these extracts, aside from their 
supposed specific secretion content, were also heterologous 
protein, a similar extract of dried sheep liver was used, as a 
control. The other details of the experiment were similar to 
those previously described. There were 12 non-operated rats 
in each group. The data of this experiment are given in table 2. 



TABLES 





raacurrAOB 


PKaCBHTAOa 

unum tauanota 


Liver 


58 
50 
25 
42 
67 


42 


Pituitary 


50 


PftiifTreai!. . , 


75 


Adrenal .... 


• 


58 


Thyroid 


33 







In the animals injected with heterologous protein derived /rom 
the adrenal and thyroid glands there was little or no change in 
the percentages of the two reaction types as compared with the 
group injected with heterologous liver protein. When pancreas 
extract was injected, however, there was a predominance of the 
minus reactions, which is in contrast to the pancreas-free ani- 
mals in which injections of heterologous protein produced a 
predominance of plus reactions. 

The fact that, after single gland removal, but one tjrpe of 
reaction occurred in certain groups led to a study of the rela- 
tionship of the endocrine chain in its effect upon the reaction 
tjrpes. Double gland ablations were done in such fashion that 
all possible two gland combinations were obtained, always 
excepting those with the pituitary. 



BLOOD SUGAB EQUILIBRIUM AND NEOPLASIA 



227 



Six rats were used in each group. At varying periods after 
operation these animals were injected with 0.5 cc. of homologous 
protein and sugar estimations were made as before. The data 
of the experiment are given in table 3. 

When compared with single gland removal, double gland 
ablation showed the following changes in type of reaction. The 
spleen and testes combination showed results opposite to those 
of spleen and ovaries, plus types predominating with the former 
and minus types with the latter. The thymus-thyroid combina- 
tion gave a predominance of plus types while the thymus-adrenal 
gave a majority of minus reactions. Three groups showed one 
type of reaction throughout; the adrenal-pancreas, testes-pan- 









TABLE 8 


















TK8TB8 


OVABIM 


BPLBKf 


THTICUB 


paucbbas 


ADSBTAL 




+ 


33 
50 
100 
40 
25 


+ 


80 
25 
67 
25 


100 
20 
75 
33 
75 


+ 

50 
50 
60 
60 


50 
50 
40 
40 


+ 

67 
32 
52 


33 

68 
48 


+ 

80 



20 
100 


+ 

40 


— 


Thjrroid 


67 
50 

60 
75 


60 


Adrenal 




Pancreas. 


. 


ThjrmuB 

Spleen 









+ " plus reaction percentage 
— ■■ minus reaction percentage 

creas, and the thjrroid-ovary combination gave 100 per cent 
minus reactions. 

As has been stated in a previous paragraph not only did nor- 
mal animals differ in type of reaction, but they also differed in 
the intensity of the blood-sugar disturbances. If the intensity 
of the reaction, as measured by the difference in milligrams per 
100 cc. of the two sugar estimations, be compared in the young 
and old (chart 1) it is found that the young react more distinctly 
than do the old. Of the two sexes the males react more 
distinctly than do the females. 

If the intensity of the reaction which follows the injection 
of heterologous protein be compared with that which follows 
the injection of homologous protein in normal and in gland-free 



ROHDENBURO, EBEHBIEL AND BERNHARD 



Ccmpdrjsaa ef hisimty of Asiurlancs of SlooiSv^i^iIibriim 
w Jottgk 01tnd-BceJium*JSwHk 7m IhffereBilfpez ^Aoiigea 



t 




BLOOD SUGAB EQinUBRIITM AND NEOPLASIA 229 

animals (chart 2), it is found that the same degree of reaction 
occurs with either antigen except in animals with the adrenals 
or the thyroid removed. The adrenal-free group reacted more 
distinctly to heterologous protein, while the thyroid-free reacted 
more intensely to homologous protein. Spleen-free animals 
showed a lessened intensity of reaction to both antigens. 

When the reaction following injection of homologous protein 
is studied in animals from which two glands have been removed, 
the results (2), to quote from the paper previously referred to, 
are as follows: ^'Tliere are two combinations of endocrines 
which when ablated inhibit the reaction; these are (1) spleen 
and either gon^d set, and (2) adrenal and pancreas. If one 
gland from either system be ablated and with it one gland from 
the other system, there results little or no change from the nor- 
mal reaction. If, however, one gland from one system be 
removed and with it one gland from another system, and in 
addition either the thyroid or thymus, then marked inhibition 
occurs. Removal of one gland from one system and either 
thyroid or thymus does not produce inhibition. If both glands 
be removed from each sjrstem, then, in order to induce inhibi- 
tion, both of the intermediate glands, i.e., thyroid and th3rmus, 
must also be removed; the ablation of one intermediate gland 
is not sufficient. From this it appears that there are two sys- 
tems which control the reaction and these systems are connected 
through two apparently interchangeable glands." 

Stimulation of the reaction with double gland ablations oc- 
curred in two systems, analogous to those producing inhibition. 
These systems were the thyroid in combination with either 
ovaries, pancreas, or spleen, and the pancreas in combination with 
either testes, thyroid, or spleen. 

The exact significance of variations in type and in intensity 
of reaction is somewhat in doubt. It might be supposed, since 
disturbances of blood-sugar equilibritmi follow the injection of 
an antigen until the injected animal has exhausted the power 
to produce further antibodies against the antigen injected, 
that variations in the type or in the intensity might be indica- 
tive of the relative strength of the antibodies formed. 

TBS JOURNAL Or CAMCER BC8BABCB, VOL. yi* NO. 8 



230 ROHDENBURG^ KBEHBIEL AND BERNHABD 

Whether such an hjrpothesis is valid or not was determined in 
the following experiment. Rabbits, in groups of four animals 
each, were prepared as follows: one group served as normal 
controls; from a second the testes were removed; while from a 
third the complete thyroid apparatus was excised. One week 
after operation and after twelve hours' starvation blood-sugar 
estimations were made, the animal was then injected intraperi- 
toneally with 2 cc. of a 50 per cent suspension of washed sheep 
red cells, and one hour after the injection the blood sugar was 
again determined. Lytic titers were made with serum obtained 
from the first bleeding. Two subsequent injections and double 
sugar determinations as well as lysin titers were made at weekly 
intervals, and a final lytic titer was made one week after the 
last injection. As is shown in chart 3 the thyroid-free animals 
showed a greater sugar disturbance than did the control animals, 
and the control animals in time showed a greater disturbance than 
did the testes-f ree animals. If intensity of reaction bears relation 
to the strength of antibody formation then we should expect 
the thyroid-free group to show the highest lytic titer, and the 
testes-free the lowest titer. Such, however, was not the case, 
for the normal controls gave the highest titer, the testes-free 
next, and the thyroid-free the lowest titer. 

There being no demonstrable relationship between the intensity 
of the reaction and the strength of the antibody development, 
the most plausible remaining explanation is that the phenom- 
enon is an indication of the relative activity of various of the 
endocrines. The failure of the injected animal to react after 
the maximum antibody production has been attained might also 
be taken as an indication of the temporary exhaustion of the 
endocrines. 

It is more logical, in endocrine studies, to consider averaged 
group behavior rather than individual behavior, and it is on 
averaged group behavior that the scheme of interrelationfihip 
given below has been worked out. If the ablation of a gland 
or of a group of glands is followed by a minus reaction, then 
that gland or the combination when in situ may be considered 
as causing a plus reaction. Similarly, if a given gland or gland 



BLOOD 6UGAB EQUILIBRIUM AND NEOPLASIA 



231 



group removal brings about an inhibition, then that particular 
combination when in situ could be considered as stimulators 
and vice versa. The deviation one way or another from the 
behavior of the control group of normals would serve in lesser 
degree to indicate further interglandular relationships. 



jfFjt GJfca4uaue Ihsiurhanca, 



— AnnJCanhels 
7hyr^d/r99 



32 




soo 

wo 

TOO 
(U 

400^ 
300^ 

/oo 




charts 



Thus as is diagrammatically shown in chart 4, if the injection 
of a homologous protein is followed by a drop in the blood-sugar 
concentration, either the thyroid or, if the individual be a male, 
the testes, may be considered as ineflScient; whereas, if the sugar 
values rise, the adrenals may be considered as overactive. It 



232 ROHDENBUBG, EREHBIEL AND BERNHARD 

may also be supposed that decreases in the sugar concentration 
are due to lessened activity of the adrenal whose function is 
to keep up the sugar concentration. This insufficiency may be 
due either primarily to the adrenal or secondarily to the in^- 
fectireness of the stimuh reaching that gland through the i^leen, 
thymus, or pancreas. In a similar manner, a rise in the blood- 
sugar concentration may be attributed to failure of the thyroid 
or testes to hold the blood sugar at the lower level, and this 



may in turn be due to the thyroid or testes primarily, or it may 
occur secondarily because of failure of the stimuli from either 
spleen, pancreas, or thymus. 

The intensity of the reaction has been worked out on similar 
lines. Accepting, for example, that a normal intensity of reac- 
tion for the rat is 31 mgm., if a gland removal results in reac- 
tion intensity of but 10 mgm. then the gland removed may be 
considered when in situ to be one of those helping to ke^ up 



BLOOD SUGAR EQUILIBRIUM AND NEOPLASIA 233 

the normal intensity. If two glands be removed from two dif- 
ferent groups and it is found that group A has a reaction intensity 
of 45 mgm. and group B a reaction intensity of 10 mgm. it may 
be accepted that the gland whose removal brought about the 
increased intensity, is when in situ one of the group which keeps 
the intensity down, the opposite being the case in the group in 
which gland removal caused a less intense reaction. Now, if 
both glands be removed from a third group and the reaction 
intensity of this group is found to be 10 mgm., it may be supposed, 
gland A in situ being an inhibitor, and gland B in situ being a 
stimulator, that gland A sends an inhibiting influence to gland 
B, for gland A being removed permits gland B to act without 
restraint, and the reaction is intensified. Gland B being re- 
moved permits gland A or some other gland to act without 
restraint in consequence of which the reaction intensity falls. 

The intensity of reaction according to glandular interrela- 
tionship is indicated in chart 5. For example, if the adrenal 
be removed, the spleen lacks a stimulating influence and the 
thyroid an inhibiting one; the retention of a normal reaction 
intensity might, therefore, be expected, for the plus and minus 
influences balance each other, the actual experimental values 
being: normal group, reaction intensity 31 mgm.; adrenal-free 
group, reaction intensity 36 mgm. Ablation of the spleen re- 
sults in the removal of secondary stimulation received via the 
gonads, adrenal, and pancreas, and the removal of a stimulatmg 
impulse sent to the pancreas, as well as directly upon the blood- 
sugar concentration. It might be expected, therefore, that the 
splenic removal would depress the intensity of the reaction. 
The actual experimental values obtained were: normal animals, 
reaction intensity 31 mgm.; spleen-free group, reaction intensity 
10 mgm. 

Theoretically, therefore, if the blood sugar concentration 
after an injection of homologous protein falls distinctly, it 
may be supposed that the mechanism which causes increased 
concentration (chart 4) has not responded or that the mechan- 
ism which depresses has been stimulated. An intense reaction 
would speak for thyroid inhibition. In contrast, if the reaction 



234 BOHDENBURO, EREHBIEL AND BERNHASD 

were slight, say 10 mgm., splenic inhibition would be suspected. 
Inversely, if the sugar concentration rises distinctly it would 
indicate adrenal hyperactivity in combination with thyroid 
and pancreas, while if the reaction were Tninimal it would indi- 
cate adrenal hyperactivity in combination with spleen and 
gonad. 

That disturbances of blood-sugar equilibrium no longer occur 
after the injection of protein, when the organism has attained 



its maximum antibody producing power is, on this basis, to be 
interpreted as indicating temporary exhaustion of certain of the 
endocrines. 



CLINICAL APPLICATION 



It has ah%ady been noted that a majority of the mice bearing 
i^MMitaneous tumors subjected to this test have shown a de- 
creased intensity of reaction. In order to follow this lead, 
further investigations were carried out on human beii^ suf- 



BLOOD SUGAB EQUILIBRIUM AND NEOPLASIA 236 

fermg from a large variety of pathological conditions. The 
technic of the test as applied to the human cases was as follows: 

After the evening meal no medication was given, the patient 
was starved until the last specimen of blood was withdrawn the 
following morning. Just before and again 45 and 120 minutes 
after the subcutaneous injection of 10 c.c. of homologous pro- 
tein, blood was withdrawn and the sugar content determined. 
The degree of disturbance of equilibrium was estimated by 
subtracting the lowest from the highest of the three values 
obtained, the difference in milligrams being the standard of 
comparison. The homologous protein used was either blood 
serum or ascitic fluid obtamed from mdividuals free from 
syphilis, tuberculosis, or cancer, in order to obviate any possible 
chance of disease transmission. The sterility of the fluid was 
assured by culture. No reactions were encountered as a result 
of such injections except in one instance where there was a slight 
chill followed by a rise in temperature to 102°F. With the 
Folin-Wu method we have found that not more than three 
unknowns can be compared with the same standard because 
the colors of the standard fade rapidly. 

The clinical diagnoses have m every instance been confirmed 
either by histological or bacteriological examination, autopsy, 
operation, or investigation of the subsequent course of the 
patient. All the cases of neoplasia have been confirmed 
microscopically. 

Before proceeding to an analysis of the data obtained as a 
result of these investigations it is profitable to revert to those 
obtained during investigations already reported (3). In this 
previous investigation blood-sugar equilibrium was disturbed 
by the oral administration of 100 grams of anhydrous glucose. 
When the individuals examined by this method were divided into 
male and female, cancerous and noncancerous, the following 
results were obtained. There was no predominance of type in 
one group over another, i.e., a particular type of reaction could 
not be said to be characteristic of any given group. If the 
intensity of the reaction, i.e., the degree to which blood-sugar 
equilibrium was disturbed was compared (chart 6) it was found 



236 



BOHDENBUBQi EREHBIBL AND BEBNHABD 



that with the exception of the non-cancerous males all the groups 
gave the greatest percentage of cases with a reaction intensity 
of from 61 to 120 mgm. ; in brief, there was nothing characteristic 
of neoplasia.' 

However, when the data of the types and intensities of reac- 
tion observed after "the injection of homologous protein in man 
were similarly analyzed there were some very interesting and 
suggestive variations. 



WTJ^m of '^pe dndlntensiiy of Dii^arbance cf Wood Sugar 
afier Glucojso hgesiion in CaMcer and Mm Cancer Aareitz 

l^ms. Z0^60tO/00/20UO/iOaOZOO 

Qfn nm-cancer cases ss% gsn plus reachdns 

a 94 cancer cases ssr^gavejalus rcachons. 

Iniensity cfHeacbon, 
f^^ Cancer Cases 

\ Non' Cancer Cases: 



1 



XOi 



IS 



10 




Cli^6 



Table 4 gives the statistical data of the 209 cases which have 
been examined. When this information is presented as in 
table 5, several points are evident; a larger percentage of the 

* In a recent publication Killian and Kast (4) stated that the h3rperglycemia 
observed in cases of neoplasia is attributable to the nitrogen retention present 
in many of these cases. This factor had not been considered by us in our previous 
communications. A review of our data, however, does not bear out the conten- 
tion advanced by these authors, but confirms in a very emphatic manner the 
observation made by Spence (5) who attributed the hyperglycemia to the age of 
the individuals. In his investigations Spence showed that hjrperglycemia occurs 
with increasing frequency in the more advanced age periods. 



BLOOD STJGAB EQinLIBRITTM AND NEOPLASIA 



237 



TABLE 4 



CAsa 

NUMBXB 



DIAOirOSU 



6432 General carcinomatosis. 

^S2740 Carcinoma rectum 

177 Carcinoma lung 

6977 Carcinoma stomach 

7068 Multiple myelomata . . « . 

71 18 Carcinoma stomach 

02 Carcinoma cecmn 

54 Carcinoma pancreas — 

262 Carcinoma rectum 

391 Carcinoma gall bladder 

528 Carcinoma breast 

794 Melanooarcinoma skin. . 

1260 Carcinoma stomach 

1542 Sarcoma lung 

1573 Carcinoma stomach .... 

1659 General carcinoma 

1735 General carcinoma 

1563 Creneral carcinoma 

1820 Sarcoma femur 

1761 Carcinoma colon 

1853 Carcinoma stomach 

1944 Carcinoma rectum 

2165 Carcinoma general 

2192 Carcinoma stomach 

2465 Carcinoma stomach 

3056 Carcinoma general 

2892 Carcinoma general 

3152 Carcinoma stomach . . . . , 

2942 Carcinoma rectum 

3752 Carcinoma stomach 

3892 Carcinoma cervix 

4647 Carcinoma stomach . . . . , 

5442 Carcinoma general 

P2386 Carcinoma stomach 

P2306 Carcinoma stomach 

P2488 Carcinoma sigmoid 

6368 Carcinoma bladder 

6479 Carcinoma stomach . . . . , 

P2820 Sarcoma ovary 

6950 Carcinoma vagina 

831 Carcinoma bladder 

1235 Carcinoma prostate 



DUTUBBAHCB 

or 

BLOOD BUOAB 

mgtn. 

-9 
-9 

+4 

-3 

+4 

-12 

-6 

-10 

+6 

+5 

+3 

+4 

-2 

+7 

+6 

-H 

+5 

+6 

+10 

-4 

-5 

+7 

+9 

-11 

+10 

+4 

+10 

-3 

+4 

-5 

+9 

+7 

+4 

-9 

+12 

+7 

+6 

+5 

-5 

+3 

-6 

+3 



238 



ROHDENBURG, KBEHBIEL AND BEBNHABD 



TABLE 4-CoiiliiMMt 




1353 
1273 
1587 
1825 
1869 
2114 
2306 
2025 
2709 
2471 
3604 
3517 
3627 
3021 
3858 
1740 
3421 
3533 
3886 
4107 
4213 
4216 

225 
4510 
4134 
4414 
3097 
14140 
4747 
4565 
4403 
4706 
5113 
4748 
2500 

128 
5193 
4531 
5821 
6044 
3225 
5508 



Carcinoma stomaoh 

Garoinoma liver 

Carcinoma stomach 

Carcinoma stomach 

Carcinoma esophagus. . . 

Carcinoma rectum , 

Carcinoma rectum , 

Carcinoma breast 

Carcinoma generalised. . 
Carcinoma generalised. . 

Carcinoma cervix 

Carcinoma breast 

Carcinoma liver 

Carcinoma esophagus. . . 

Carcinoma colon , 

Carcinoma esophagus. . . 

Sarcoma parotid 

Sarcoma pelvis 

Sarcoma retroperitoneal 

Carcinoma stomach 

Carcinoma breast 

Carcinoma rectum 

Carcinoma stomach 

Carcinoma stomach.^ . . . 

Carcinoma stomach 

Epithelioma cheek 

Carcinoma stomach 

Carcinoma stomach 

Carcinoma stomach 

Carcinoma stomach 

Carcinoma stomach 

Carcinoma huynx , 

Carcinoma stomach 

Carcinoma cervix , 

Carcinoma intestine. . . . . 

Carcinoma liver 

Carcinoma cervix 

Carcinoma tongue , 

Carcinoma breast 

Carcinoma colon , 

Carcinoma breast 

Carcinoma sigmoid 



or 

BLOOD SUOAB 

-5 

+3 

-10 

-10 

-6 

-6 

+10 

-8 

-10 

-2 

+4 

-6 

+9 

+6 

+6 

-6 

+4 

+2 

+4 

-8 

-5 

+8 

+2 

-6 

+4 

-4 

+10 

+6 

+3 

-6 

-2 

-4 

+3 

+4 

+12 

-6 

-12 

-10 

+12 

+9 

+10 

+8 



BLOOD SUGAR EQUILIBRIUM AND NEOPLASIA 



239 



TABLE 4— Conltiiiieii 



CAMM 



5886 

5889 

158 

6153 

754 

700 

2300 

2815 

3605 



80 
4086 
128 
5637 
5793 
3728 
5746 
6116 
6049 
6347 
1807 
1757 
2768 
3202 
3560 

4965 

594 

797 

864 

835 

1102 

1219 

1379 

770 

1580 

1373 

1299 

2222 

2662 

2641 

30 



DIAOHOUB 



Carcinoma stomach 

Carcinoma breast 

Carcinoma stomach 

Carcinoma e8ophag;u8. . . . 
Retroperitoneal sarcoma 
Carcinoma esophagus — 

Carcinoma stomach 

Carcinoma Imig 

Carcinoma esophagus. . . . 
Sarcoma tendon sheath. . 

Carcinoma colon 

Carcinoma stomach 

Epithelioma penis 

Carcinoma uterus 

Carcinoma uterus 

Lymphosarcoma neck 

General carcinoma 

Carcinoma ovary 

Carcinoma bladder 

Carcinoma mouth 

Carcinoma general 

Carcinoma appendix 

Carcinoma uterus 

Carcinoma testes 

Carcinoma stomach 

Carcinoma lip *. 

Endothelioma brain 

Syphilis 

Inguinal hernia 

Encephalitis 

Biliary cirrhosis 

Encephalitis 

Syphilis 

Neuritis 

Tuberculosis 

Achylia gastrica 

Gastric ulcer 

Orchitis 

Varicose veins 

Leukemia 

Fibroadenoma breast 

Cryptorchidism 



DXSTURBAlId 

or 

BLOOD fUGAB 



mom. 

-3 

-9 

+6 

+6 

-33 

-35 

-51 

-51 

+54 

-f20 

+36 

-34 

-42 

-27 

+30 

+28 

-34 

-33 

-31 

+30 

-21 

+19 

+22 

-20 

+49 

+14 

+30 

-48 

-44 

-24 

-15 

-16 

+21 

+42 

+15 

+15 

-21 

+15 

+18 

-42 

+18 

-39 



240 



ROHDENBX7BO, KBEHBIEL AND BERNHABD 



TABLE 4--C<mttmMtf 



CAS* 
KUMBBB 



DXAOMOSIS 



DnrruBBAJfCB 
or 

BLOOD aUQAB 



31 
32 
36 
37 
3209 
3889 
3822 
4464 
2467 
4792 
4825 
4937 
4943 
5121 
126 
5244 
5197 
5356 
5312 
5407 
5452 
5319 
5645 
5582 
5628 
5654 
5830 
5842 
5814 
6115 
6046 
5406 
6530 
6603 
6841 
237 
3183 
527 
583 
100 
1465 
1109 



Cryptorchidism 

Angioneurotic edema. . . . 

Hyperthyroidism 

Retinitis pigmentosa — 

Pernicious anemia , 

Pernicious anemia 

Orchitis 

Ovarian cyst 

Endocarditis 

Myocarditis 

Cardio-nephritis 

L3rmphadenitis 

Fibromyoma uteri 

Endocarditis 

Gastritis 

Postoperatiye adhesions 

Pregnancy 

Hypertrophied prostate . 

Fistula in ano 

Catarrhal jaundice 

Graves' disease 

Pernicious anemia 

Subacute gastritis 

Catarrhal jaundice 

Endometritis 

Oophoritis 

Neurasthenia 

Gastric ulcer , 

Splenomegaly , 

Cholelithiasis 

Mixed timior parotid. . . . 

Cerebral thrombosis 

Diabetes 

Uterine fibroids 

Papilloma bladder 

Cholelithiasis 

Diverticulitis 

Gastric ulcer 

Neurasthenia 

Tuberculosis , 

Ovarian cyst 

Cholecystitis 



-20 
-18 
+28 
+45 
+28 
+36 
-32 
+29 
+21 
+28 
+48 
+62 
-40 
+35 
-86 
-18 
-18 
+34 
+16 
+39 
-33 
-22 
-24 
-17 
-39 
+24 
-17 
+20 
-63 
-22 
+23 
-53 
+60 
+45 
+22 
-42 
-15 
-28 
+26 
+16 
+15 
+33 



BLOOD SUGAR EQT7ILIBRIX7M AND NEOPLASIA 



241 



TABLE ir-Condvd^ 



CAta 

KaUBMM 



1639 
1604 
1809 
1827 

478 
2554 
2588 
2689 
8229 
5436 
6843 
5664 
1511 
35 
5058 
5372 
6401 

141 
5565 
5727 
6016 
6063 
6943 
7123 

136 

198 
7086 

984 
1008 

369 
1306 
1873 
2293 
3121 
3351 
3665 
3205 
4440 
4421 
5570 

201 
3666 



DZAOWOSIB 



Pyelitis 

CholecyBtitis 

Sciatica 

Chronic inflammation. . . . 

Gastric ulcer 

Tuberculosis 

Chronic cholecystitis 

Gastric neurosis 

Cholecystitis 

Angina pectoris 

Gastric ulcer 

Lung abscess 

Pernicious anemia 

Pregnancy 

Gastric ulcer 

Cystooele 

Duodenal ulcer 

Orchitis 

Syphilis 

Cholecystitis 

Duodenal ulcer 

Pernicious anemia 

Chronic appendicitis 

Nephritis 

Bronchopneumonia 

Fibroid uterus 

Fissure in ano 

Prolapse uterus 

Dirertictditis 

Duodenal ulcer 

Fibroid uterus 

Nephritis 

Cholecystitis 

Chronic endocervicitis. . . 

Cholecystitis 

Tuberculosis 

Hepatitis 

Cholelithiasis 

Polyp of sigmoid 

Hypertrophy of prostate 

Septic endocarditis 

Pernicious anemia 



DlflTUBBAMCa 

OF 
BLOOD SnOAB 

mom, 

+22 

+18 

-15 

-14 

+16 

+20 

+18 

+15 

-17 

-39 

-38 

-15 

-9 

+6 

+4 

+10 

+6 

+10 

+10 

+ 

+10 

+9 

-7 

+11 

-8 

+10 

+2 

+1 
+10 

+2 
+2 
-9 
+5 
+4 
+3 
+4 
-7 
-5 
+3 
+7 
+2 
+9 



242 



BOHDENBUBOy KBBHBISL AND BBRNHABD 



cases of neoplasia of the gastro-intestinal tract and of the liver 
and gall-bladder show reactions of an intensity of less than 12 
mgm., than is the case in individuals who have non-malignant 
diseases of these organs. In diseases of the genito-urinary 
apparatus this relationship does not apparently hold true, but 
in the series of miscellaneous neoplasias, no one type bdng 
present in sufficient numbers to warrant separate tabulation 
it is again evident that a larger proportion of the cancerous 
individuals show a lessened intensity of reaction than do a 
similar group of non-cancerous individuals. 

TABLES 



Neoplasia of genitourinary tract 

Neoplasia of gastrointestinal tract 

Non-grouped neoplasia 

Non-malignant disease of genitourinary tract 

Non-malignant disease of gastrointestinal tract . . 
Non-grouped non-neoplastic diseases 

AU cases of neoplasia 

AU cases non-neoplastic 



EBACnmr 



Leas than 
12mgin. 



7(64%) 
«0 P8%) 
29 (74%) 

9 (41%) 
18 (38%) 

9 (20%) 

86 (78%) 
81 (81%) 



uoro UMn 
12 



6 (46%) 

7 (12%) 
10 (26%) 
13 (69%) 
21 (62%) 
35 (80%) 

23 (22%) 
60 (60%) 



These differ^aces are distinct when compared with the simi- 
larity of behavior found in our previous series where blood-sugar 
eqiulibriima was disturbed by the ingestion of glucose. The 
observation is possibly of value as a diagnostic procedure. In 
diseases involving the gastro*intestinal tract, from mouth to 
anus, duodenal ulcer excepted, and in diseases involving the 
gall-bladder and liver, 88 per cent of all the cases of neoplasia, 
Le., 50 of 57 cases, showed a reaction of less than 12 mgm. As 
regards the reaction types it was found that 39 per cent of the 
neoplasia cases gave plus reactions as compared with 52 per cent 
in the non-cancerous group. If the previously outlined endo- 
crine relationship be accepted, such lessened reaction would 
indicate either spleen, adrenal, or gonad insufficiency, or thyroid, 
adrenal, and pancreas overactivity. 



BLOOD BUGAB EQUILIBRIUM AND NEOPLASIA 243 

CONCLUSIONS 

1. The disturbance of blood sugar equUibrium which foUows 
the injection of a protein varies in intensity, and in type. 

2. The removal of certain of the endocrines influences both 
the intensity of the reaction and its type. 

3. Neither the intensity of the reaction nor the type have 
any relation to the strength of the antibody development after 
the injection of a selected protein. 

4. A study of the type and reaction intensity has been used 
in an attempt to establish paths of endocrine correlation. 

5. Human cases of neoplasia show a larger proportion of 
weak sugar reactions, i.e., of 12 mgm. or less, than do non- 
cancerous individuals. This is particularly true of neoplasms 
of the gastro-intestinal tract and liver. 

REFERENCES 

(1) RoHDBNBUBG AND Pohlman: Am. JouT. Med. Sol., 1020, clix, 853. 

(2) RoHDXNBXTBO AND Kbbhbixl: Am. Jour. Med. Sci., 1^1, clzii, 28. 

(3) RoBDENBUBQ, Bbbnhabd, AND Kbbhbibl: Am. Jour. Med. Sci., 1020, 

clix, 577. 

(4) KiLLiAN AND Ea8t: Arcli. Int. Med., 1021, xxviii, 813. 

(5) Spbncb: Quart. Jour. Med., 1021, xiv, 314. 



PREGNANCY AND TUMOR GROWTH 

ISIDOR KRCN9S 
From Cohmbia Vnivernty, IrutittUe of Cancer R$9earch, F. C. Wood, Director 

Received for publication January 16, 1022 

The relationship between pregnancy and tumor development 
has been investigated by a nmnbpr of observers, whose results 
have been, however, in many instances contradictory. For 
example, Moreau (1) and Herzog (2) found that transplanted 
tumors grow with unusual rapidity in pregnant animals. Haaland 
(3), on the other hand, maintained that pregnancy exerts a 
markedly restraining influence on the growth of transplanted 
tumors, which continue to develop in their umial fashion aftet 
parturition. Bashford and Murray (4) hold that "pregnancy" 
in the female ''and full sexual activity in the male constitute no 
bar to successful transplantation." Uhlenhuth and Weidanss 
(5) observed retardation in the growth of tumors during 
pregnancy and also more frequent spontaneous retrogression. 
Bridr6 (6) noticed a low percentage of positive inoculations in 
pregnant animals, and Ehrlich (7) observed that inoculations 
into animals bearing yoimg was frequently followed by negative 
results and that in those successfully inoculated the growth of 
the tumor was greatly retarded. Then, again, Albrecht and 
Hecht (8) maintained that pregnancy affects the growth of 
tumors just as little as the presence of a tumor influences con^ 
ception or gravidity. Cuenot and Mercier (9) reported a most 
interesting observation. They noticed that Borrel's tumor "B," 
which rarely underwent spontaneous absorption, if inoculated 
before fecimdation developed during gestation and receded dur« 
ing lactation. The tumor, however, did not regress if one mouse 
only was bom, thus leaving the activity of the mammary gland at 
a minimum. Neither did absorption occur even in the presence 
of several young, if the tumor was so situated that its vascularis 

245 



246 ISIBOB KROSS 

zation was independent of that of the mammary gland. Fichera 
(10) explained these various inconsiste^ncies by assmning that 
when many embryos were present the food stuff was almost wholly 
consimied by them, while if there were only a few some of this 
material was available for the timior cells. 

Slye has concluded from her observations of spontaneous 
tiunor growth in the mouse that: First, reproducing females 
grow much less tumor than do non-reproducing females of the 
same age, etc. Second, reproducing females grow much less 
tmnor while they are reproductive than they do while they are 
non-reproductive. In other words, she finds that pregnancy 
exerts a markedly retarding influence upon growth of tumor. 

In re-investigating this problem rats were employed. The ani- 
mals selected were young adults, sexually mature, vigorous, and in 
very good condition, which had been separated from male animals 
for a period of one month to avoid the possibility of imrecognized 
pregnancies. The rats were mated with strong, vigorous young 
males that were kept in the same boxes not only during the entire 
period of gestation but also after the females had littered, in 
order to impregnate the same animals immediately after par- 
turition. In all cases the date of conception was calculated by 
coimting back twenty-one days before the date of littering. All 
the animals were inoculated on the same day with 0.003 gram of 
Flexner's rat carcinoma, the time chosen being two days after 
the animals had been mated. In all, 105 animals were inoculated. 
Of these, 43 (41 per cent) became pregnant and littered normally. 

In analyzing the cases in which no growth occurred, it is interest- 
ing to note that about 7.8 per cent of the males inoculated showed 
no growth, and that about 8.4 per cent of the females showed no 
growth. In the group of females without tumor growth, 44 per 
cent did not become gravid. The other 46 per cent did. These 
figures tend to show that in the negative cases neither sex nor 
gravidity played any r61e. 

Tracings of the tumors were made every four days, and in 
this way growth was depicted simply but effectively. 

A comparison of the charts (figs. 1 to 4) of the tumors of the 
pregnant and of the non-pregnant groups shows at a glance 



PREGNANCY AND TUMOB OBOWTH 247 

that in these experiments pregnancy had absolutely no effect at 
all upon the growth of the transplantable Flexner rat carcinoma. 



niGHhHT CONTROL, 

• - • • • • o • 



% • " 



• • • • • f 

• • t 



• • • • 












• • • t • • 






-f 






— — — • • 



• •••«# 

*• LITT£^£0. ^ 



10 e«. •• •••• 

Fig. 1 Fig. 2 

DISCUSSION 

It seems impossible completely to reconcile all of the conflicting 
results which have been published in regard to the effect of 
pregnancy on the growth rate of inoculated tumors, chiefly be- 
cause we have insuf&eient data upon which to base any conclusions 



348 



ISIDOB KSOSS 



as to the growth capacity of the tumors which the various 
experimenters have employed. Nor do we know, in many in- 
stances, the site of inoculation. Mice are especially unsuitable 
animals for such a study, as the implanted tumors are apt to be 



FKlQtHAHT 



• • • • 



CONTROL. 






# # 



o o 



• 






• • • 



• # • # • 

• « • • « • t 

• • • • f 1 1 



• * • 



• • t • • 



• I * 



. . » 






• • • • 



: • • • t 
• • • • • • 

ji 

: • f t • • 

X'LITT£feQ 



• • t 



# i *# 



• • • 

• # t 

• • • 



# • f 



• 



•• • • 



» • 



• # • 



Fig. 3 



IttM. 

Fig. 4 



variable in their growth rate, and may become so large as to kill 
the animal quickly. This obviously complicates the problem 
and renders careful control and statistical analysis of the results 
necessary. But it does seem possible to draw a few conclusions. 



PREQNANCT AND TUMOB GROWTH 249 



not entirely inharmonious with the reported experimental work 
and the facts obtained by clinical observation. 

The timior grafts in the experiments here reported were not 
placed in the mammary gland but by intent in the axilla; they 
were, therefore, not affected by the increase in the vascularity of 
the breast at the beginning of pregnancy as they would be if 
placed directly in the main mass of the mamma. This increase 
in vascularity is undoubtedly the explanation of the rapid growth 
of tumors in the breast observed in human beings. But the 
conditions in the human female and in female mice are not 
comparable. A mouse weighing 15 to 20 grams wiU frequently 
carry a tumor weighing one-t^ith of its weight. Unfortunately, 
the weights of the tumor-bearing animals are not given in Miss 
Slye's paper, though the dimensions of the tumors are furnished. 
On the other hand, the weight of the tumor in a human female 
is rarely more than one one*hundredth or one one-hundred and 
fiftieth of the body weij^t. Mammary cancer in a mouse, there- 
fore, may correspond in mass to a rapidly growing breast car- 
cinoma of 10 to 15 pounds in a human female, a tumor which is 
practically never seen. It is quite possible that the demand for 
food for such a theoretical tumor if complicated by pregnancy 
mi^t result in the starvation of the tumor for a time as Moresohi 
(12) has shown to be posmble. 

It is, therefore, evident that Miss 81ye's results, while applica- 
ble to one species of mammalia under the conditions which she 
specifies, are by no means obviously transferable even to other 
types of rodents, much less to human beings. 

The fact that in the additional experiments here reported no 
influence on the growth rate of tiunors was shown is probably 
due to the disproportion between the size of the rat and its em- 
bryos, frequently weighing 150 grams with 25 grams of embryos, 
and the tumors, which are usually not more than one-fiftieth 
to one-himdredth of the weiglht of the animal. So, too, in 
mice the proportionate weight of the embryos demanding food 
is larger than that ordinarily occurring in the human female, 
a child rarely being over one-fifteenth of the mother's weiglht, 
while in the multiple mouse pregnancies 5 grams of embryos may 
be produced at a single birth from a mother weighing only 22 
grams. 



250 ISIDOR KROSS 

A grafted tumor, such as we have employed, being a mass of 
somatic cells growing in a foreign host, affords a possibility for 
studying the problem of tissue growth independently of organ or 
other relationshix)s. The timior is simply implanted in the sub- 
cutaneous connective tissue, while a primary timior of the breast, 
in its inception at least, has important anatomical correlations 
with the tissue in which it grows. It is thinkable that some of 
the results observed by Miss Slye may be due to changes in the 
connective tissue stroma of the ttmaor rather than to the epithe- 
lium, but in the implanted tumor the connective tissue is furnished 
by the host, the epithelium being derived from another animal, 
so that the situation is quite different. 

It seems safe to conclude, therefore, from our own escpm- 
ments, that pregnancy of itself does not necessarily alter timaor 
growth rates, but that such interference, when it is observed, is 
due to the forced division of food substance between the tumor 
and the offspring; similar checking of the tumor can be observed 
in implanted tumors when the food of the non-i»^gnant animal 
is reduced approximately to the starvation point. 

A comparable diminution in the tumor growth rate may be 
seen in human beings in the terminal stages of extreme cachexia 
with innutrition but, as a rule, in man pregnancy eHher has no 
effect on the progress of a cancer or, if the mammary or uterine 
tissued are involved, hastens the growth. 

There is nothing, therefore, in the varying results of published 
experimental work which cannot be harmonized or which con- 
troverts the clinical observations already recorded. 

REFERENCES 

(1) MoRBAu: Arch, de mM. exp6r. et d'an&t. path., 1894, vi, 008. 
^) Hsrzog: Jour. Med. Res., 1902, xi.8. iii, 76. 

(3) Haaland: Berl. klin. Wchnschr., 1907, xxiii, 718. 

(4) Bashford and Murray: First Sci. Rep., Imperial Cancer Research Fund, 

London, 1904, p. 14. 
(6) TJhlsnhuth and Weidanz: Arb. a. d. k. Gsndbtsamte., 1909, xxx, 440. 
(6) BaiDRi: Ann. de I'Inst. Pasteur, 1907, xxxi, 763. 
GT) Ehrlich: Verhandl. d. deutsch. path. Gesellsch., 1908, xii, 29. 

(8) Albrbcht and Hecht: Wien. klin. Wchnschr., 1909, xxii, 1783. 

(9) Cuxnot and Mbrgisr: Compt. rend. Acad. d. so., 1909, exlix, 1012. 

(10) Fighbba: Jour. Exper. M., 1911, xiv, 320. 

(11) Sltb: Jour. Cancer Res., 1920, v, 25. 

(12) MoRBSOHi: Ztschr. f. Immunitfttsforsch., Orig., 1909, ii, 651. 



IS CANCER MORTALITY INCREASING? 

WENDELL M. STRONG 
Received for publication March 17, 1022 

To determine whether or not there is a real increase in cancer 
mortality we must compare cancer death rates for the same ages, 
since cancer is an old age disease. The effect of age distribution 
is so great that it is safe to say that any considerable increase in 
the proportion of the population at the older ages, for instance, 
the emigration of the young men and women, would cause a 
noticeable increase in the cancer deaths per 100,000 population, 
while an influx of young men and women would decrease them. 

Dr. Frederick L. Hoffman, statistician of the Prudential Life 
Insurance Company, in a paper on "The Menace of Cancer" 
published in 1913, and in his comprehensive book on "The Mor- 
tality from Cancer" published in 1915, has given the statistics 
available at the time of writing in a very complete form. These 
statistics were drawn both from the United States and abroad. 
From these I have selected a few. 

In the registration of 1900, the cancer death rate per 100,000 
population, which will hereafter be spoken of as the cancer death 
rate or cancer mortality, increased from 63 in 1900 to 79 in 1913, 
an increase of over 25 per cent. England and Wales for the same 
period showed an increase from 83 to 105 and most other Euro- 
pean states showed a very considerable increase. Going back 
forty years, the rate in twenty large American cities was 49 for 
the period 1881-1885; for 1913 it had increased to 89. In twelve 
European countries the rate was 69 for the period 1896-1900; 
for the period 1906-1910, it had increased to 81. These were 
general population statistics without subdivision by ages. There 
were also available, however, some extensive statistics subdivided 
by ages. The cancer death rate in the registration states of the 
United States of 1900 subdivided into ten-year age groups, 25-34, 

261 



252 WENDELL M. STRONG 

35-44, etc., for 1901 compared with 1911 shows an increase in 
the rate for each age group for the later year, such increase being 
very considerable for ages above 45, which are the significant 
ages in cancer. Massachusetts, for the two periods 1901-1905 
and 1906-1910, with a different distribution into age groups and 
subdivided by sexes, showed an increase for each adult group and 
for each sex separately, in the later period the increase again 
being very considerable for the older groups. While these are 
but a few of the statistics, they are illustrative of the trend of 
practically all, and, standing alone, would appear to indicate a 
great and rapid increase in cancer mortality. We have, however, 
later and very extensive statistics from two of the great insurance 
companies which are not in accord with the above. 

In 1919, Dr. Louis I. Dublin, statistician of the Metropolitan 
life Insurance Company, published in this Jottbnal (1) the 
experience of that company on its millions of industrial policy 
holders for the years 1911-1916. Tables 1 and 2 are from this 
experience. They show fluctuations but no definite tr^d. 

Table 1 is for the age group 55-64 only, which group I>. Dublin 
chose as illustrative. We find in it the lowest mortality for white 
males in 1912 and 1915, the highest in 1913 and 1914; for white 
females the lowest in 1911 and 1914, the highest in 1912 and 1915. 
It is a curious accident that the years of low mortality for the one 
sex are in three cases the high for the other. 

Table 2 compares the first two years of the period with the last 
two and in total shows but one per cent difference between them, 
a difference smaller than would be e3q)ected from merely acciden- 
tal fluctuations. The two features which may be of significance 
are that the total of males shows an increase of 5 per cent and 
that the differences in each age group, as well as in the total of 
all ages (except perhaps 65-69), are no more than would be ex- 
pected from accidental fluctuations. 

The cancer experience for the Mutual life Insurance Company 
for the years 1915-1920 has just been complied by Dr. Brandreth 
Symonds, chief medical director. This experience, given in 
table 3 below, is not subdivided by sexes, but as the great pre- 
ponderance of insurance was on white male lives it may be taken 
as representing white male cancer mortality. 



18 CANCBR MORTAUTT INCBEA8IKG? 



253 



TABLE 1 

Metropolitan Life Indueirial Department. Cancer death rate per 100,000, 

agea 65 to 64 







WHITB 


COLOBXO 


TBAB 


AILCLAMM 










Main 


Females 


Hake 


Ftomalee 


1916 


386.4 


368.0 


427.4 


218.3 


339.9 


1915 


380.8 


336.0 


427.8 


175.7 


394.3 


1914 


300.9 


385.0 


423.3 


167.7 


361.7 


1913 


384.1 


370.3 


414.6 


195.2 


368.3 


1912 


881.9 


334.1 


443.2 


176.4 


325.4 


1911 


368.7 


363.3 


400.2 


158.0 


373.7 



TABUBS 

Metropolitan Life Indueiridl Department. Ratio of cancer death rate for yeare 
t9il^t916 to death rate for yeare 191t-19t$ by percentagee 







ws 


m 


OOMIIBD 


AQB FnUOD 


AUsCLueaa 










Ifalei 


Fttneles 


Make 


Femake 




ptrcent 


p€re9nt 


ptretni 


pere$ni 


peretnl 


25 and over 


101.0 


105.2 


98.3 


106.1 


100.7 


25 to 34 


98.0 


121.0 


94.1 


95.5 


95.0 


35 to 44 


100.1 


104.3 


96.9 


142.6 


96.2 


45 to 54 


99.1 


109.7 


97.0 


67.7 


96.1 


55 to 64 


102.2 


101.1 


101.2 


118.0 


106.2 


65 to 74 


107.0 


115.9 


101.9 


117.2 


107.0 


75 and over 


101.0 


104.1 


104.4 


110.7 


48.6 



TABLES 

Mutwd Life Ineurance Company. Cancer death rate per 100,000 



▲OB FBBZOD 


19U 


1016 


1917 


1918 


1919 


1920 


AXXTBABa 

19U-1920 


25-29 


15.34 


3.96 


0.90 


7.53 


1.47 


5.08 


5.35 


80-34 


11.52 


15.19 


19.55 


10.54 


5.52 


5.49 


10.55 


35-39 


7.61 


12.95 


13.38 


10.99 


5.74 


18.69 


11.87 


40-44 


31.39 


45.31 


30.86 


26.92 


33.73 


21.21 


31.07 


45^9 


40.36 


41.82 


51.59 


60 96 


51.56 


78.80 


55.15 


50-54 


166.47 


157.36 


100.67 


79.26 


73.45 


70.02 


106.78 


55^ 


132.22 


213.84 


204.03 


153.07 


152.73 


150.95 


168.88 


60-64 


235.74 


366.73 


255.24 


308.53 


285.56 


272.73 


287.21 


65-^ 


225.77 


380.94 


532.84 


357.91 


502.42 


423.88 


407.99 


70 and over 


554.09 


1000.67 


834.83 


856.57 


731.75 


982.20 


829.54 



254 WENDELL M. STRONG 

If the first two years, 1916-1916, combined are compared with 
the last two, 1919-1920, we find that up to age 65 the cancer 
mortality ran in general considerably lower for the later than for 
the earlier period, while for ages 65 and older the reverse was 
true. Of the nine five-year groups (including ''70 and over" as 
such a group), beginning with age 30-34, five show a higher mor- 
tality in the earlier two-year period and four in the later. Had 
the subdivision into age groups been the ten-year groups of the 
Metropolitan, beginning with 25-34, every group up to age 65 
would evidently have shown a higher mortality in the earlier 
period. 

The year 1915, for some unexplained reason, was a year 
of very low cancer mortality in the Mutual, a peculiarity 
which did not appear in the Metropolitan experience and 
which was, therefore, probably accidental. It is consequently 
worth while to see what would have been the result had 1915 
been eliminated so that we should have compared the years 
1916-1917 with 1919-1920. We find that of the nine five-year 
groups, seven would have shown a higher mortaUty for the earlier 
period and only two for the later; moreover, had the grouping 
been by ten-year groups, as in the Metropolitan experience, each 
ten-year group from age 25 up would evidently have shown a 
higher mortality in the earlier period than in the later. 

Again, a comparison of the last two years combined with the 
preceding four years combined, shows that of the nine groups 
five had a higher mortality in the earlier period; we must recog- 
nize, however, that two of the four were 66^9 and ''70 and over" 
which are of greater importance than groups near the lower age 
limit. Had the exceptional year 1915 been eliminated, however, 
the result would have again looked extremely favorable for the 
later period. 

In order to have the most recent possible data, the approxi- 
mate cancer mortality for the year 1921 has been worked out, 
although exact figures are not yet available. The 1921 results 
compared with the average for the preceding six years show a 
lower mortality for each of the seven five-year age groups from 
30-34 to 60-64, but a higher mortality for 65-^9 and "70 and 
over." 



IB CANCSB MOBTAUOT INCREASINO? 255 

Emphasis is given to the failure of the Mutual experience to 
show any tendency to an increasing death rate by the fact that 
the lives involved were nearly all white males, and in other ex- 
periences where the sexes are separated any tendency towards 
increase has generally been more among males than females. 

The Metropolitan experience has been extended from 1915 
through 1020 but has not yet been made public. Through the 
courtesy of Dr. Dublin I have had an opportunity to see these 
statistics and to include in this paper the deductions from them. 
They show in general the same lack of any tendency to increase 
in the age groups from age 30 to age 65, but they show an mcreas- 
ing tendency from age 65 upwards. 

ThuS; we have had the experience of the two companies cover- 
ing the eleven-year period from 1911 to 1921, inclusive. The data 
in these experiences should probably be at least as accurate as 
to the causes of death as those of any other experience, because 
the insurance companies try to obtain accurate knowledge of 
the cause of death at the time the claim is paid. The figures 
indicate, taken at their face value without considering whether 
there are any modifying influences, for ages below 65 either a 
fluctuating or a slightly decreasing cancer mortality; for ages 65 
and over they vary, but probably, on the whole, indicate some 
increase. 

The above are the direct deductions from the statistics. Be- 
fore accepting these as final results, however, we must go back of 
the statistics and inquire whether there are any circumstances 
or considerations relating to the data on which these statistics 
are based tending to modify the results. One such consideration 
should be mentioned. With the development of medical science 
there has been a gradual increase in correctness of diagnosis of 
the cause of death. Professor Walter F. Willcox (2) showed in 
1917 that a very considerable apparent increase in cancer mor- 
tality would result simply from the continued improvement 
in the correctness of diagnosis, which increase in correctness had 
probably continued up to the present day. Undoubtedly in the 
past many deaths attributed to old age should properly have 
been set down as cancer, and many others attributed to other 
causes would undoubtedly have been attributed to cancer if the 
correct diagnosis had been made. The result of this would be 



256 WENDELL M. 8TBONO 

that if there were a really stationary cancer mortality it would, 
nevertheless, appear to be increasing considerably because of the 
increasing correctness of diagnosis. 

How effective this could be to cause an apparent increase 
where no real one exists may be seen in the statistics of appendi'* 
citis mortality which show an increase of 40 per cent from 1900 to 
1015; yet undoubtedly the real rate was decreasing in this period 
because of surgical advance and readiness to resort to surgery* 

Moreover, increasing correctness in diagnosis works almost 
altogether in one way; tiiat is, it results in attributing to cancer 
many deaths which would formerly have been incorrectly attrib- 
uted to some other cause, and in changing but few the other 
way. The effect of increased correctness in diagnosis, which is 
imdoubtedly more important at the advanced ages, would seem to 
me enough to explain what apparent increase in cancer mortality 
there is in ages above 65, and it would emphasise the fact that 
for ages below 65 there has not been even an apparent increase. 

Meanwhile the statistics of the United States Registration 
Area show a continuous increase in cancer deaths. These are 
population statistics, however, and are not analyzed as to ages; 
hence, as compared to such statistics as we have considered, they 
have very little meaning, since a change in the age proportion 
of the population might be responsible for the entire apparent 
cancer increase. 

My conclusion would be, even taking into account the regis* 
tration area's apparent increase, that we cannot now determine 
whether the cancer mortality is slightly increasing, practically 
stationary, or slightly decreasing, but that we can be sure it is 
not greatly increasing. A more exact result is something for 
future investigations when reliable statistics for a long period of 
years are obtainable. 

Lest what I have said be misinterpreted I would add that such 
a conclusion does not lessen at all the seriousness of the cancer 
problem. It merely holds out hope that the terrible scourge will 
not increase without limit. 

REFERENCES 

(1) Dublin, L. I: Mortality Statistics of Cancer Among Wage Earners, etc., J. 

Cancer Res., 1919, iy, 235. 
C2) Wnxcox, W. F: On the Alleged Increase of Cancer, Trs. Am. Statistical 

Assoc, 1917, XT, 701 and J. Cancer Res., 1917, ii, 267. 



CANCER AND PARASITE 

ISIDOR KROSS^ 
From Coluwhia Univeraiiy, Inaiituie of Cancer Reaeareh, F. C. Wood, Direetor 

Reoeiyed for publication May 1, 1922 

Ever since the first recognition of the bacteria as a cause of 
disease, attempts have been made to demonstrate a bacterial 
cause for malignant tumors. The importance of such a dis- 
covery, with all its possibilities of prophylaxis or even cure, is 
obvious, and accounts for the sustained interest in this question, 
in the face of constantly repeated failure. 

One of the most recent investigators is Nuzima (1) who em- 
ployed a mouse carcinoma (no. 11) from the Crocker Institute. 
From this growth he isolated a diplococcus and described a 
malignant tumor having all the characteristics of carcinoma 11, 
which followed inoculation of a culture of this organism into mice. 

The importance of such a discovery woxdd be, needless to say, 
beyond computation, and it was therefore d^tc^miiieid to repeat 
the experiments with additional checks and controls. That 
microorganisms are frequently found in malignant txmiors, espe- 
cially those growing in mice, is a well established fact, but, as 
has been said, no one has ever succeeded in proving that they 
play any r61e as causative agents. 

To summarize Nuzum's findings, it will suffice to say that he 
cultivated pieces of tiunor in tissue ascitic fluid media covered by 
paraffin; and thus partially anaerobic. From this culture he 
isolated a Gram-positive diplococcus, which he considers the 
specific bacterium in the production of carcinoma 11. A sub- 
culture was made of this organism in the same manner, and 
inoculated into mice. In three of them, the typical carcinoma 11 

^ Adjunct Gynecologist, Mt. Sinai Hospital, New York. 

257 

TKB JOUBNAL OF CAMCBB BMBABCH, TOL. TI, NO. 4 



258 ISIDOB KROSS 

is said to have developed. However, careful perusal of the article 
suggests that only two of these loice had a real tumor. 

As can readily be seen, two factors must be considered in an 
experiment of this nature: (1) the tumor cell proper; and (2) the 
microorganism, if any. In order to test the ability of the latter 
to produce the maUgnant growth m question, it is necessary to 
prevent any participation of the tiunor cells in the process. To 
do this without affecting the microorganism, two methods were 
employed in my own experiments, radiation and freezing. 

In the first, several mice that had well developed carcinoma 11 
tumors, were exposed to the Roentgen ray, being given a dose 
known to destroy the tiunor cells completely. The rayed tumor 
was then excised aseptically and inoculated into 24 mice in the 
usual fashion. Part of this same tumor was cultured, following 
most minutely the technic laid down by Nuzum. As a control, 
carcinoma of the same series, from another mouse, was inocu- 
lated into 24 mice, and fragments were put at the same time 
into twelve tubes of the ascitic tumor fluid media. The tumor 
used in this control series was not radiated. 

Transplantation of the unradiated control timior resulted in 
the usual 60 per cent of takes which this tumor has given during 
the past seven or eight years at the Crocker Laboratory. On 
the other hand, in not a single instance did a tumor develop 
from the radiated cells. In other words, where the tumor-cell 
factor was eliminated the microorganism factor, which was being 
tested, was not able to reproduce the tumor. 

Because of the extreme importance of the question, this experi- 
ment was repeated in another series of mice, with the same nega- 
tive results. 

The cultures made from the radiated t\unor were incubated 
for five days and were then inoculated into 60 mice. In none 
of the 60 did a tumor subsequently develop. At the same time, 
smears were made from the bottom of the culture tubes and 
stained with Gram's and with Giemsa's stain; these contained in 
most instances many short and long bacilli, streptococci, staphylo- 
cocci, and diplococci. In some smears all these varieties were 
present and most of the smears, in fact, showed more than one 



CANCER AND PARASITE 259 

organism. In no instance was a pure culture of any one organ- 
ism found. 

The second method employed to elimmate the tumor ceillfi 
was alternate freezing and thawing of the tumor before inocu- 
lation. The growth was removed aseptically and placed in a 
sterile dish, where it was emulsified with scissors, and the dish 
was then placed in an ice and salt mixture. After about fifteen 
minutes, when the mass was frozen solid, it was ground with a 
sterile pestle until it became soft and fluid. This freezing and 
thawing was repeated twice, with the purpose of diBstroying the 
tumor cells without harming any bacteria that might be present. 
Part of this emulsion was inoculated into 36 mice, and another 
part was cultured in the ascitic tissue fluid media. 

In only two of the 36 mice was there any growth, and in both 
instances the tumor was not discoverable until the twenty- 
eighth day after inoculation, whereas in routine laboratory inocu- 
lation, growth is generally noted in from seven to ten days. 
Hence the freezing and thawing had killed all but a few of the 
tumor cells. If carcinoma 11 were due to a microorganism, all, 
or at least most of the mice, shoidd have had tiunors, both in 
the frozen and thawed tumor series, in the rayed tumor series, 
and in the culture inoculation series. The absence of a timior 
in all but two late instances, militates against the assumption 
that carcinoma no. 11 can be ascribed to a microorganism, espe- 
cially in view of the fact that no one type of organism was found 
as a constant occurrence. The two growths that did develop 
from frozen and ground tumor can be readily explained by the 
assumption that the freezing did not destroy quite all the timior 
cells. Others of the staff at the Crocker Laboratory have had 
the same experience. 

In all the cultures the results, as concerns the specific organ- 
ism described by Nuzum, were negative as has been already 
stated and subcultures inoculated into 24 mice produced no 
tumors. 

A careful perusal of the experiments of Nuzimi leads to the 
conclusion that in only two in3tances out of eighty-nine did the 
culture fluid inoculation cause tiunor growth, i.e., animal no. 



260 ISIDOR KBOSS 

250 in experiment 1, and a second animal in experiment 3. 
These two instances and our own negative findings repeatedly 
and carefully checked up suggest that in all probability the 
tumors in Nuzum's experiments were spontaneous new growths. 
It may be stated in conclusion that all the animals showing 
negative results were kept under observation for a period of 
three and a half months, so that ample time was given for a 
timior to arise. 

CONCLUSIONS 

1. While most of the ascitic tissue fluid cultures of Crocks 
Institute carcinoma 11 contained microorganisms of various 
kinds, in no instance was there found the characteristic micro- 
organism described by Nuzimi. 

2. In not a single instance was the inoculation of mice with 
these cultures followed by tumor growth. 

3. It is probable that the two undoubted tumors in Nuzum's 
series were a spontaneous new growth. 

REFERENCE 
(1) Nuzttm: Surg. Gynec. and Obst., 1921, zxxiii, 167. 



TERATOMAS AND THEIR RELATION TO AGE 

H. E. HIMWICH 
From Department of Pathology, Cornell Unweraity Medical College, New York 

Received for publication May 10, 1922 

The great accumulation of reported cases of teratomas offers 
an excellent opportunity to review the data, with the object of 
investigating the relation of their frequency to the age of the 
host. In the course of the study, it has become apparent that 
this relationship is so definite as to assume the form of a gen- 
eral law. 

The largest collections of teratomas were gathered by TaruflB 
and Ahlfeld. Both of these observers stressed the large con- 
genital forms which are situated in the head, the thoracic, the 
abdominal, or the sacral regions. Gonadal teratomas occur 
most frequently in early adult life (Wilms). Among others, 
Askanazy investigated the internal craniopagi, and Ekehom, 
the internal thoracopagi. Lexer and Nakayama studied the 
abdominal inclusions and pygopagi. 

Though many hypotheses have been advanced on the origin 
of teratomas, they may be resolved into two view points: The 
teratoma is either the offspring or the twin of its host. Stockard 
has recently produced experimental evidence in favor of the 
latter conception. 

CRITERIA AND METHODS OF STUDY 

Two precautions have been observed in compiling the present 
statistics: (1) Only growths of tridermal or bidermal origin 
have been considered. No such case has been omitted. (2) 
Special effort has been exerted to determine the age at which 
the teratoma began its growth. The first increase in size of 
external growths may be accurately observed. For those 

261 



262 H. E. HDCWICH 

situated internally, the initial symptom was used as an indicator. 
Where the history was deficient, the age at which the operation 
took place or at which death occurred was taken as the closest 
approximation obtainable. The last criterium particularly 
applies to the teratomas of the aged. In this manner 895 cases 
have been studied. Sometimes a period of slow growth is 
followed by one of heightened activity. Such is the case of 
chorioma testis reported by Jackson, in which growth commenced 
at the age of twenty and slowly continued to twenty-three, 
after which the increase in size became extremely rapid. Since 
the relationship between the growth of the host and that of the 
teratoma is of interest, in such cases, the beginnings of both 
periods have been noted. A similar effort was made for internal 
teratomas, thus bringing the total number of growths tabulated 
to 975. 

In sjnstematizing the results it was noted that though the 
variation from year to year is considerable, there seems to be 
an orderly waxing and waning of the number of cases to an 
extent which justffies the drawing of a curve. An average has 
been drawn in order to minimize accidental variation. Six 
year periods have been chosen because they are the longest 
which correspond to actual changes throughout the length of 
the curve. The first period begins at fertilization and ends at 
five. In the curves which are drawn to a scale of one-half, 
the abscissae represent the age of the host when the tumor began 
its growth, the ordinates the number of cases in each year. 

THE CX7HVE FOR ALL TERATOMAS 

After an initial maximmn rise, the curve falls and remains 
low from five to eleven years (fig. 1). At eleven it achieves a 
higher level, which is increased at seventeen and twenty-three 
years. The second maximum is found between twenty-three 
and twenty-nine years. The curve falls gradually at twenty-nine 
and thirty-five years, and then more rapidly at forty-one, after 
which it becomes progressively lower towards its end at seventy- 
six years. Eighty-seven per cent of the teratomas occur before 
forty-one years and 95 per cent before fifty-three. 



TERATOMAS AND THEIR RELATION TO AGE 



263 



Most congenital teratomas do not evince postnatal growth and 
since this study concerns only those which do grow^ for the early 
maximal total may be substituted the smaller number of tumors, 
showing power for growth, represented by the lower broken 
line in the graph. With this correction the highest point in the 
curve is found between twenty-three and twenty-nine years. 
This is borne out by Wilms who finds the period of greatest 
frequency for sex-gland teratomas to be between the ages of 
twenty and thirty years. His conclusion is to be expected 
because teratomas occur most frequently in sex glands. 




FXQ. 1. CXTBVX FOB ALL TXBATOMAS 

The absciasae stand for the years of appearance of growths, the ordinates for 
the number in each year. Both are drawn to a scale of one-half. The total 
number of congenital tumors is not shown. The dash line indicates the average 
for each six-year period. In the first six years there are two; the upper one 
stands for all cases, the lower for those which had power of postnatal growth. 
The mode is from twenty-three to twenty-nine, the time when growth stops. 



STUDY OP THE CORRECTED CURVE 

There are three aspects of the described phenomenon: First, 
it is evident that the total number of teratomas at any age 
increases with the actual growth of the individual. As size 



264 H. E. HIMWICH 

increases, the total number of teratomas increases. But it 
must be remembered that while size is increasing, growth rate 
is falling. Therefore, second, teratomas become more frequent 
as growth slows down, at the time when growth potential be- 
comes smaller. Growth must be recognized as involving two 
elements: increasing actual proportions, and decreasing growth 
potential. Hence, third, the total niunber of teratomas at any 
age increases as growth potential diminishes. 

When one recalls the fact that the changes in growth rate 
are not constant, the number of teratomas is seen to bear even a 
closer relationship to the growth of the hosts than has been 
indicated. Not only do these tumors appear as growth of the 
host slows, but during their time of appearance the teratomas 
are more frequent in the periods of slower growth of the host. 
Teratomas are contunon in early infancy following a space of the 
most rapid proliferation of all — ^fetal growth. They increase 
again when the comparatively rapid growth rate of early child- 
hood gives way to the slower one of pubescence and lastly they 
are found in greater nmnbers as active growth gradually ceases. 

To explain the relative munber of teratomas appearing in the 
several periods we must take into consideration an additional 
factor, the growth potential of the embryonal rest, for a teratoma 
by most theories arises in an embryonal rest of some kind. It 
is known that the great majority of embryonal rests do not grow; 
they either degenerate or remain dormant. Others achieve a 
more or less perfect adult growth, while a few develop into tumors. 
Tridermal rests act in a similar manner. Thus we have a few 
with high growth potential, many with less capacity for growth, 
and finally others which remain latent unless they are stirred 
to development by an external stimulus. 

During the first six years of life the growth rate of the host 
declines rapidly, bis growth potential is greatly reduced, and 
the number of tei^tomas of relatively higher potential, capable 
of proliferation, is comparatively large. From five to eleven 
years the growth rate is fairly constant, the loss of potential is 
small, and the additional number of teratomas released is few. 
The next appreciable change takes place during the differentia- 



TEBATOMAS AND THEIB RELATION TO AGE 265 

tion occurring at puberty and accompanying the maturation 
of the sex organs. Here the number of teratomas begins to 
increase and continues to do so until growth finaUy stops, at 
which time the greatest number of teratomas make their pres- 
ence known. Although the loss of growth potential in the host 
becomes smaller, being least in the iSnal period from twenty- 
three to twenty-nine years, yet just because growth itself is 
slow, there is an ever increasing number of embryonal rests of 
low potential, capable of expressing their latent growth energy. 
The very inactive rests eyen at this period do not have suiBEicient 
energy to start growth spontaneously and are therefore consid- 
ered in another group at a later time. However, they may serve 
as a nidus for neoplasms since many do commence development 
after the growth of the host has stopped. 

Thus the number of teratomas appearing at any given time 
depends upon the amount of loss in growth potential of the 
hosts during that time and the niunber of rests whose potential 
is large enough to proliferate imder these conditions. Evidently 
tile number of teratomas in any period varies inversely to the growth 
potential of the hosts and direcUy as that of the embryonal rests. 

ANALYSIS OF CURVES OF AGE INCIDENCE OF TERATOMAS IN THE 

DIFFERENT LOCATIONS 

The object of this study is to show the relation to the general 
law of the occurrence of teratomas in the various situations. 

Thoracopagi. Ekehom in his collection of teratomas of the 
anterior mediastinimi finds fifteen cases occurring between the 
ages of twenty and thirty, four between thirty and forty years, 
and four more to sixty, thus agreeing with the present curve in 
showing the maximum occurrence from twenty-three years to 
twenty-nine. 

Abdominal incliLsions. The abdominal tridermal growths 
are frequently discovered at birth but continue to be found 
throughout life, with a second rise from twenty-three to twenty- 
nine years. 

Craniopagi and pygopagi. There is not a sufficiently large 
nimiber of head and sacral teratomas to yield reliable statistics. 



266 



H. E« HIMWICH 



SUUistica of various authors. They show a close agreement with original staiistice 
here presented. Note increased number of dermoids during pubescence 



TBOBA 


COPAOI 
nXAN) 


DBSIIOIDS 
(PAUU) 


■1IBBTOMB0 
(CBBTAMV) 


BBIfXHOIfBB 

(CHBVABBU) 


CABCnrOltAB 


VCRJiU 




Tortia 


OvBiy 


10-20 


1 


1-6 


4 


0-6 


6 


0-19 





0-6 


2 


11 


20-30 


19 


6-10 


3 


6-17 





20-26 


1 


6-11 


2 


26 


30-40 


7 


10-16 


10 


18-24 


16 


26-30 


8 


11-17 


1 


48 


40-60 


2 


16-20 


8 


26-29 


11 


30-^ 


8 


17-23 


12 


19 


60 + 


2 


20-26 


12 


30-34 


11 


36-40 


19 


23-29 


29 


26 






26-30 


14 


36-39 


7 


4046 


14 


29-36 


30 


21 






30-36 


10 


4044 


3 


46-60 


4 


36^1 


34 


16 






36-^ 


11 


46^9 


4 


60-60 


2 


41-47 


27 


16 






40-16. 


8 


60-69 


2 


60-63 


2 


47-63 


10 


14 






46n60 


10 


60-76 


2 






6a-s» 


8 


14 






60-66 


6 










69-66 


1 


1 






66-60 


3 










66-71 


1 








60-66 


1 




















66-70 


2 




















70 + 


2 


















31 




103 




61 




69 




167 


207 



Inclusions according to age and situation. Those which are capable of growth appear 
in greater number from twenty-three years to twenty-nine years 



AOB 


GBAmaPAGX 


mORAO- 
OPAOI 


ABDOMIHAL 
PABA8ITBB 


OTABIAX 
TBBATOMAB 


TBBATOMAB 


PTQOPAOX 


TOTAL 


-6 


78 


66 


36 


17 


48 


46 


279 


6-11 


1 


3 


6 


16 


7 


1 


33 


11-17 


4 


11 


18 


27 


9 


1 


70 


17-23 


4 


14 


16 


44 


32 


4 


114 


2a-29 


1 


16 


20 


62 


66 




144 


29-36 




10 


16 


47 


44 




116 


3&-41 




4 


8 


39 


43 




94 


41-47 




4 


4 


22 


16 




46 


47-63 




4 


4 


19 


7 


2 


36 


63-69 




1 


1 


13 




' 


10 


69-66 




1 


2 


10 


3 




16 


66-71 






1 


6 


1 




8 


71-77 


2 




1 


2 






6 


Total 


90 


123 


132 


313 


264 


63 


976 



TERATOMAS AND THEIR RELATION TO AGE 267 

Most are congenital. The internal craniopagi occur most 
frequently from eleven to twenty-seven years. The only year 
where there is more than one case is the nineteenth, where 
there are two. The last teratoma capable of spontaneous growth 
appeared at twenty-seven. The later examples of sacral tera- 
tomas are even rarer than those occurring in the head. There 
is one each at ten and thirteen years and four from nineteen to 
twenty-three. 

Ovarian teratomas. Turning next to the gonadal teratomas 
and comparing the testicular and ovarian curves, we see that 
the latter is less variable (fig. 3). This is due to the delayed 
diagnosis of so many of the ovarian teratomas, probably because 
of their slow growth, as their structure is often of the adult type. 
Their internal position further postpones their discovery, yet 
dermoids are sometimes found by accident. Nevertheless, 
the largest number of tumors occurs between the ages of twenty- 
three and twenty-nine. From a review of one himdred and 
three cases Pauh finds dermoids appearing most often from 
twenty to thirty years. 

Another difiference between the ovarian and testicular curves 
is the greater rise in the former in the two periods between 
eleven and twenty-three years. This phenomenon might be 
expected as a result of the growth differences since, in the male, 
postpubescent growth is more rapid than in the female. 

Teratoma testis. The external situation of the male sex gland 
allows prompt discovery of its tumors. Although the greatest 
decline in rate of growth occurs early in life from birth to four 
years and the nimiber of testicular teratomas at this time is 
large, the maximimi nimiber nevertheless occurs at a later period. 
This is probably due to the fact that growth under four years is 
comparatively very rapid in spite of its fast declining rate. The 
modal year of the curve is twenty-six. It is interesting to note that 
the last growth cartilage of the long bones ossifies at twenty-five. 

Teratomas in the testes commence growth most frequently 
between the ages of twenty-three and twenty-nine years. Che- 
vassu finds the maximiun between twenty and thirty years 
(fig. 2). 



268 H. E* HIMWICH 



TEBATOMAS OF LATER LIFE 



After establishing the time of greatest frequency of teratomas, 
it still remains to account for those of old age. Those diagnosed 
after the age of fifty-one may be divided into two groups. The 
first consists of neoplasms of adult structure which had reached 
the limits of their capacity for post-natal growth, while those of 
the second are more malignant. To the first group belong such 
tiunors as the following: 

Craniopagi. Beck reports a case in which a dermoid was 
found at autopsy in place of the hypophysis in a woman seventy- 
four years of age. Eberth reports a similar accidental finding 
beneath the diu'a in a woman of seventy-five. 

Thoracopagi. There are two examples in Ekehorn's collec- 
tion — ^Finder's case of a patient with bulbar paralysis, aged fifty- 
three, in whom the dermoid was discovered at autopsy; and 
Lebert's of a man of sixty who had been dyspneic since his six- 
teenth year. 

Abdominal incltLsions. Rizzoli (Tarufii) reports two cases 
of late abdominal inclusions, one at sixty, the other at sixty-two. 
Symptoms had been present for a long time in both. In one 
of the cases they appeared first at the age of twenty. 

The tumors mentioned thus far were benign, though some 
produced symptoms because of their size and position. 

In the second group of neoplasms the element of trauma be- 
comes important in the etiology. There is Bonney's report of a 
retroperitoneal chorioma of a man of sixty-seven, and Goebell's 
of an abdominal teratoma that became malignant at fifty-four, 
twenty-seven years after a mass had been diagnosed. Djewitski 
reports a chorioma of the bladder which first gave symptoms 
at the age of seventy-three. The same irritation which produces 
a papilloma of the bladder may transform an otherwise benign 
embryonal rest. 

Pygopagi. Hudson describes a sacral teratoma which began 
growth at the age of fifty-two; the history shows that a nodule 
had existed in that region since the birth of the patient. The 
histological picture is one of a tridermal rest with cancerous 
degeneration of mucous glands. It ia similar to an old age 



TERATOBiAS AND THEIB RELATION TO AGE 269 

cancer arising in previously normal tissue. Evidently in the 
last four cases it is not growth potential of the embryonal rest 
but an extrinsic traiunatic influence^ to which every part of the 
body is subject^ that caused the proliferation of cells. A case 
of Briddon's beautifully Ulustrates both these factors occurring 
in the same growth but independently and at different times. It 
concerns a sacral dermoid which appeared externally at the age 
of twenty-two and then ceased growth till the fifty-second year, 
when it underwent epitheliomatous change. 

Teratoma testis. There are three examples of late teratoma 
testis. Lexer quotes one from v. Bergmann's clinic in a man of 
sixty. On section the growth was of adult structmre. 

Ewing and Fepere report cases which first showed growth at 
the ages of sixty-one and sixty-three, respectively. The micro- 
scopic examination in both instances showed carcinomatous 
change of one element in a totipotent rest. 

Ovarian teratomas. Of thirty-two dermoids, thirteen exhibited 
malignant transformation of a carcinomatous, sarcomatous, or 
endotheliomatous tjrpe. Four showed thyroid structure, of 
which three were rapidly growing tumors. In six reports de- 
tails were lacking. However, since the growths were called 
dermoids, their structure must have been of the adult type, 
like that of the remaining nine inclusions. 

To smnmarize, the late appearance of stationary teratomas is 
due to their delayed discovery, while that of growing teratomas 
is caused by their injury. 

Carcinoma testis. Carcinoma testis is discussed in this place 
not only for its possible teratomatous origin, but because trau- 
matic etiology links it with the tumors of later life. In many cases 
of teratoma testis in yoimg people the transformation of a slowly 
into a more rapidly growing tumor is caused by trauma. In 
older people the growth is rapid from the start. The same se- 
quence of events obtains for carcinoma testis. The cell of 
many cases of carcinoma testis is characteristic, with a large 
nucleus and clear cytoplasm. Sometimes the growth is called 
a sarcoma; the difference in opinion is due to the fact that no 
analogous cell is foimd in the human body. This same cell is 



270 



H. £. HIMWICH 



often found with teratoma testis. There are only two probable 
interpretations: (1) The irritation caused by some extrinsic 
factor, in this special case, by the teratoma on the tubule cells, 
is the cause of carcinoma. (2) The unique type of cell is of 
teratomatous origin. Chevassu takes the position that it 
develops from the adult spermatogonia, putting the tumor in 
the class of acquired carcinomas. The final convincing link 
in the chain of evidence has not been produced, for he has not 
been able to trace the steps of anaplastic change from the sperma- 
togonia to the carcinoma cell. 

It has been definitely established that the cell which is of more 
rapid growth will often ovemm and may finally crowd out alto- 
gether the other constituents of the tumor. Thus arise the 
rhabdomyomas, the chondromas — ^the simple tumors of the 
sex glands. In this imcontrolled competition the most embryonal 
type of cell would have a decided advantage. Therefore Ewing 
concludes that carcinoma testis is a one-sided teratoma. In 
the light of the foregoing it is interesting to see to which of these 
two theories the carcinoma curve lends itself. 

According to Chevassu's statistics embryomas occur with 
greatest frequency from twenty-five to thirty years and semino- 
mas from thirty-five to forty (fig. 2). The writer's review of a 
larger nuimber of cases coincides with the data of Chevassu, the 
modes occuring from twenty-three to twenty-nine years in the 
teratomas and thirty-five to forty-one years in the carcinomas. 

Comparing the carcinoma testis curve with that of all cancer 
(Hoffman^), of which the congenital cases are too small a propor- 

^ Mortality from cancer throughout the United States Registration Area. All 
organs and all parts. 1903-1912. (Hoffman, The Mortality from Cancer.) 





MALI 


FEMALB 


Until 10 


1,170 


984 


10-24 


2,028 


1,844 


2&-34 


3,757 


7,891 


35-44 


10,750 


26,779 


45-54 


24,431 


46,669 


55-64 


35,327 


52,393 


65-74 


33,745 


43,010 


75 and over 


18,381 


24,601 



TERATOMAS AND THEIR RELATION TO AGE 



271 



tion materially to alter the general outline, we see that the former 
has no resemblance to the latter, for in that case it would have 
a continuous rise to some time after sixty. In brief, the carcinoma 
testis curve is the teratoma testis ^urve with the mode slightly 
shifted. 

Since the histogenesis of carcinoma testis has not been traced 
from either embryonal or adult tubule cells, it is probable that 
carcinoma testis is of nontesticular origin, and since there is no 
reason why misplaced cells should so often be of the same tjrpe 
or occur so frequently with teratomas, unless they are of tera- 




jtjt 



»»-»f 



Fio. 2. Testiculab Teratomas 



The lower two curves are reproduced from Chevassu's paper. The con- 
tinuous Ime represents the teratomas; the dash line, the carcinomas. In order 
better to compare the new curves with those of Chevassu, the number of tera- 
tomas in each year was divided by four, and that of carcinomas by two. Both 
curves are plotted on a basis of ten-year periods. 

tomatous origin, we are forced to this conclusion towards which 
the study of the curve gives additional evidence. Similarly to 
teratoma testis, carcinoma may show a congenital increase in 
the size of the organ. Morestin (Chevassu) reports such a tumor, 
which assumed malignancy at the age of thirty-seven. Like- 



272 H. £. HIMWICH 

wise carcinoma occurs more often in undescended testicles. It 
begins its growth in the rete and, like teratoma testis, is occa- 
sionally observed in pseudohermaphrodites. 

Ovarian tumors. The origin of primary tumors of the ovary 
is so imdecided that any data in reference to them is of particular 
interest. Here we shall mention a few facts in regard to one of 
these tumors which may be of teratomatous origin, i.e., the 
sarcoma. 

Ewing divides these sarcomas into three main types: (1) 
spindle cell; (2) roimd cell; (3) myxoma cell. This classification 
is of special significance since just such types of sarcomatous de- 
generation of dermoids have been observed (Debucy). 



T" 




Fio. 3. Oyabian Teratomas 

These curves are averages for six year periods drawn to a scale of one-half. 
The continuous line represents the teratomas with the mode from twenty-three 
to twenty-nine years; the dash line the carcinomas with the mode from eleven to 
seventeen years, the period of pubescence. 

Desurmont finds that the different primary tumors are bilateral 
to var3dng degrees. However, sarcomas, 25 per cent, and der- 
moids, 20 per cent (Pauli), approximate each other quite closely. 

Finally, the most common tumor of infancy is the sarcoma, 
which is most frequent at fifteen years (Donhauser). Cordier 
and Zangemeister give fifteen and twenty years, respectively, 
as the age of most common occurrence of sarcomas. They are 
found from fifteen years to twenty-five and from forty years to 
fifty (Desurmont), both periods of phjrsiological stimulation. 
The writer finds the mode of the combined carcinoma and sar- 
coma curve at fifteen years (fig. 3). Comparing the mode of 
this curve with that of ovarian teratomas, we find that it has 
been shifted forward to the time of pubescence. Hence there is 
a group of embryonal cell tumors having an age incidence simi- 
lar to teratomas, and becoming malignant under the stimulation 
of puberty. 



TEBATOMAS AND THEIB RELATION TO AGE 273 
FACTORS IN THE ETIOLOGY OF TERATOMAS 

The growth of the earlier teratomas may be adequately ex- 
plained on the basis of a growth competition between the host 
and the embryonal rest. But even in the teratomas of infancy 
another factor, traiima, may be present. It becomes increasingly 
important later on. 

Growth potential. The growth of the host inhibits that of the 
teratoma. 

The growth of the embryonal rest may be divided into 
two parts : its prenatal development, or growth which continues 
until stopped by the inhibition produced by the excessive 
growth of the host; and the growth of which it is still 
capable (growth potential) after that of the host slows down 
or ceases. These two parts are in reciprocal relation to each 
other. The earlier the prenatal inhibition, the smaller and less 
differentiated will be the inhibited rest but the greater will be 
the remaining growth potential. Small embryonal rests may 
develop proliferative powers while large ones, which achieve a 
certain intrauterine development, seldom if ever show further 
capacity for spontaneous growth. 

Trauma. The shift in the mode of the curve of carcinoma 
testis from twenty-three to twenty-nine years to thirty-five to 
forty-one years is due to an external stimulus. 

In spite of the larger growth potential of the smaller testicular 
rests many do not achieve malignancy until their immediate 
region is traumatized. This is illustrated on comparing these 
tumors in the testicle and the ovary. A larger proportion of 
the latter are benign, forming adult structures. The chief 
difference in their histories is due to their locations; the testicle 
is exposed to injury, the ovary is not. It is generally admitted 
that teratomas do not become malignant much oftener than do 
normal tissues. 

In the case of carcinoma testis we should expect with a history 
of injury to a small imdifferentiated rest and the resulting pro- 
liferation of an embryonal cell, a shift in the mode of the curve 
towards that of old age cancer. 

ma lOUBNAL OP OAlfOSB BUXABCH, VOL. TI, HO. 4 



274 H. B. HIMWICH 

StJMMABY AND C0NCLT7SI0KS 

1. Growth potential and teratomas 

1. Parasite. Teratomas are tridermal embryonal rests en- 
dowed with a certain amount of possible growth, i.e., growth 
potential. When the rest is comparatively large it has necessarily 
consumed considerable growth energy before birth, while in 
the small teratoma, the growth period may be divided into two 
parts, a slight early growth soon followed by an inhibition, and a 
later, or post-natal growth, should conditions permit. 

2. Host. The total number of teratomas in a population, 
up to any given age, increases while growth potential of the 
hosts decreases. As the larger increases in the number of tera- 
tomas occur in periods when growth of the host is slowed most, 
the growth of the host must inhibit that of the embryonal rest. 

3. The number of teratomas appearing in any given time 
varies inversely with the growth potential of the host and directly 
as that of the embryonal rest. The tumors which begin their 
postnatal growth before that of the host stops are of highest 
potential, but are not necessarily more malignant, for they must 
overcome a still present inhibition. Since most teratomas have 
a low growth potential, they appear most commonly at the 
time the growth of the host stops — from twenty-three to twenty- 
nine years. 

//. Trauma and autonomous growths 

1. Teratomas which start growth as a result of injury are 
malignant more frequently than those which proliferate solely 
under the influence of growth potential. 

2. When trauma precipitates growth, the teratoma is fre- 
quently monodermal. If the inclusion is still in an undifferen- 
tiated condition the cell is often of an embryonal tjrpe. If a 
developed inclusion is traiunatized the cell in many instances 
is like that of acquired cancer. 

3. The ciu*ve of carcinoma testis rises and falls in a manner 
similar to that of teratoma testis and not like that of old age 



TERATOMAS AND THEIB RELATION TO AGE 275 

cancer. This is another fact which may be adduced in support 
of the theory that carcinoma testis is a one-sided teratoma. In 
the female, a similar neoplasm might be expected to arise as the 
result of the physiologic stimulations of puberty. This is what 
actually takes place, hence the growth, in all probability, is of 
teratomatous origin. 

4. Trauma is followed by proliferation of cells, and any pre- 
cipitant of regeneration may be important in the etiology of 
acquired cancer. In the old the inhibition of the organism is 
almost negligible. Hence trauma at that time may readily be 
followed by an imcontrolled and therefore excessive growth. 
Thus, loss of growth restraint may be almost as important a 
factor in the etiology of acquired cancer as in that of congenital 
inclusions. 

REFERENCES* 

Babbon: Teratomata of the brain, Jour. Cancer Res., 1916, i, 311. 
Chbistian: Dermoid cysts and teratomata of the anterior mediastinum, Jour. 

Med. Res., 1902, yii, 54. 
Dbsubmont: fitude anatomo-clinique des tumeurs solides bilaterales des ovares, 

Thdse de Paris, 1911-1912. 
EwiNo: Neoplastic Diseases, Philadelphia, 1919. 

Ewinq: Teratoma testis and its derivatives, Surg., Gynec. & Obst., 1911, xii, 230. 
Lszxb: Ueber teratoide Geschwulste in der Bauchhohle und deren Operation, 

Arch. f. klin. Chir., 1900, bci, 648. 
Naxayama: Ueber Congenitale Sacraltumoren, Arch. f. Entwcklngsmechn. d. 

Organ., 1905, xix, 475. 
Stocxabd: Developmental rate and structural expression, Am. Jour. Anat., 

1920-1921, xxviii, 115. 
Tabuffi: Storia della Teratologia, 1881-1894. 

' These references were chosen out of over one hundred and sixty; they contain 
extensive reviews of the literature consulted. 



AN ATYPICAL ADENOMA OF THE PANCREAS ORIGI- 
NATING IN ISLET TISSUE 

HARRY GOLDBLATT 

From the Department of Pathology, School of Medicine, Western Reserve UnipersUy, 

Cleveland, Ohio 

Received for publication May 13, 1022 

True simple adenomata of the pancreas arising from the 
islands of Langerhans have been described by several authors. 
The condition is very rare and the diagnosis is somewhat difficult 
to establish. The tumors hitherto described have been very 
small, the largest being 11 mm. in diameter. Most of them were 
accidental autopsy findings, and some were discovered only in 
the sections prepared for routine microscopical examination. 
Nichols (10), in 1902, was one of the first to report such a tumor. 
He stated clearly and convincingly upon what groimds he based 
the diagnosis, and subsequent authors seem to have guided 
themselves according to those facts in deciding the islet origin 
of an adenoma of the pancreas as opposed to the other two pos- 
sible sources, viz., duct and adult acini. Helmholz (3) in 1907, 
Morse (8) in 1908, and Koch (6) in 1914, described cases which 
fulfilled the requirements stated by Nichols and concluded that 
they were dealing with a simple adenoma originating in the 
islands of Langerhans. LeComte (7) in 1913, described a case 
of his own and also reviewed the literature to that date. He 
determined that adenomata of the pancreas have no great clinical 
significance and ''constitute no very serious hindrance to long 
life.'' He classified as true adenomata of the islands those 
reported by Nichols (10), Ssobolew (13), Reitmann (11), Herx- 
heimer (4), Hehnholz (3), Morse (8), Cecil (1), Weichsel- 
baum (14), RoUet (12), and Heiberg (2). Although he favors a 
diagnosis of adenoma in the above cases, some of the authors 
themselves, notably Ssobolew, Herxheimer, Weichselbaum, and 

277 



278 HABBT GOLDBLATT 

Cecil, either consider the nodules to be hjrpertrophic islands, or 
do not decide definitely between a diagnosis of simple hyper- 
trophy and neoplasm. Whenever there is doubt between the 
diagnosis of simple hypertrophy and new growth he recommends 
the latter if the nodule exceeds the arbitrary diameter of 1 mm., 
provided, of course, that it shows all the characteristics of island 
tissue. Others have reported the same lesion and several have 
gone so far as to trace the origin of some carcinomata of the 
pancreas to the islands. The conclusions of some of these 
authors, especially of Fabozzi, who reported five cases, have 
been disputed vigorously. With the exception of Horgan (5) 
most of these investigators studied advanced carcinomata so 
that a doubt of validity of their conclusions seems justifiable. 

The tumor of the pancreas to be described here was discovered 
at autopsy by Dr. H. T. Karsner, performed through the courtesy 
of Dr. John Maclachlan. 

ST7HMABT OF CLINICAL HISTORY AND PHYSICAL EXAMINATION 

The patient is an unmarried white female fifty-five years of 
age, in whose past history nothing of importance is noted save 
that eight years ago a myomectomy was performed with un- 
complicated recovery. Two years ago she noticed looseness of 
the bowels, which over several months progressively became 
worse and ultimately became a profound diarrhea. This was 
accompanied by gradual loss of flesh and strength but she was 
not confined to bed until about two months before death. In 
the last weeks of her life bowel movements were extremely 
frequent but well controlled, sometimes numbering 16 to 20 in 
the course of twenty-four hours. No microscopic or chemical 
examination of the stools was made but the gross examination 
showed thin watery stools with very little mucus and no blood. 
At various times indigestion of the different food elements was 
noted. On a milk diet the stools, upon standing, showed a 
supernatant layer of undigested fat. When meat was intit)duced 
into the diet, undigested fragments were seen in the stools. 
Fruits such as banana were easily recognizable. Fermention was 
never marked nor was putrefaction severe. The clinical exami- 



ATTPICAL ADENOMA OF PANCREAS 279 

nation showed no abnormality of lungs, heart, abdomen, or 
nervous system. The patient finally died apparently as the 
result of profound asthenia. 

SUMMABY OF AUTOPSY FINDINGS 

The body is that of a white female apparently fifty-five years 
of age, showing moderate emaciation. Upon examination of the 
internal organs there were found, in addition to the tumor of the 
pancreas described in detail below, slight chronic interstitial 
nephritis, fatty metamorphosis of the liver, subacute entero- 
coKtis, and slight passive congestion of lungs, liver, kidneys, 
and intestines. 

SPECIAL EXAMINATION 

Gross description 

When the abdomen is opened the pancreas is found in the 
normal location, but a tumor mass partly embedded in the head 
presents above the lesser curvature of the stomach. This tumor 
is not adherent to any of the structures in the neighborhood. 
The pancreas measures 15 cm. in length. Partly embedded in 
the head of it, 10 cm. from the tip of the tail, is a mass, roughly 
spherical in shape and measuring 4.5 by 3.5 by 2.5 cm. This 
tumor is definitely and completely encapsulated so that it can 
be enucleated easily from the head of the pancreas. The capsule 
is thin, pearly white in color, and in it, coursing over the surface 
of the growth, are several dilated vessels, apparently veins. The 
tumor is firm but elastic and cuts with moderate resistance. 
The cut surface bulges moderately, is light yellow in color, and 
bleeds slightly. There is moderate vascularization of the tissue 
but no areas of hemorrhage are seen. Near the center of the 
mass there is one small area of necrosis about 3 mm. in diameter. 
The central portion of the growth is of a slightly deeper yellow 
color, is somewhat firmer, and bulges to a greater extent than 
the tissue at the periphery. The tissue is only moderately 
friable. The head of the pancreas is normal save for moderate 
congestion. The entire portion of the pancreas from the tumor 
to the tip of the tail is the seat of marked atrophy. In this 



280 HABRT GOLDBLATT 

portion the cross section measures 12 by 5 mm. The lobules 
are obviously reduced in size, there is slight increase of the 
interlobular connective tissue, and a moderate amount of fat 
infiltration. 

HISTOLOGICAL DESCRIPTION 

Sections were stained by hematoxylin and eosin, eosin meth- 
ylene blue, thionin, and Mallory's connective tissue stain. 

Pancreas 

The duodenal portion shows very advanced autolytic changes, 
so advanced, in fact that Uttle description is justified beyond 
noting a slight increase in connective tissue both between and 
within the lobules. Several definite islets are found which are 
of normal size and structure. The tail shows great reduction 
in the ^e of the lobules, acini, and individual cells comprising 
them. There is a great increase of the interlobular as well as 
interacinar connective tissue, and in it are a few foci of Ijrmphoid 
cells. The smaller ducts are well preserved and normal, but 
the larger ducts show desquamation of the lining epithelium in 
some instances and complete digestion in others. The islets of 
Langerhans are moderately abundant, of normal size and struc- 
ture, and fairly well preserved. There is no sign of invasion of 
the pancreatic tissue by new growth or of neoplastic change in 
the cells of the pancreas. 

Tumor 

There is a definite but thin capsule which consists of dense, 
moderately nucleated fibrous connective tissue. This shows 
no sign of invasion by the parenchyma of the tumor. Through- 
out the tumor there are present anastomosing bands of con- 
nective tissue, some of which are densely fibrous and some rather 
loosely arranged. Many of these trabeculae do not join at all 
and are seen as isolated masseis of connective tissue which vary 
considerably in size and in shape, some being elliptical or rounded, 
and others stellate or irregularly branching. Nearly every one 
of them contains one or more blood-vessels. Some of these 



ATYPICAL ADENOMA OF PANCREAS 281 

vessels are very thin-walled, consisting apparently of only a 
single layer of endothelial cells, while others are large and well 
foraied with a thick wall consisting of intima, media, and adven- 
titia. Most of the vessels are filled with blood. In some of the 
trabeculae there are seen merely blood spaces showing no definite 
endothelial Uning. The trabeculae give to the parenchyma of 
the tumor an alveolated appearance. The parenchyma con- 
sists of numerous masses and anastomosing bands and strands 
of epithelial cells. These masses vary greatly In size and shape. 



I'll;. 1. Photouicbouraph Showikg Type of Cell, Arkangement of Cbi.l.s 

AND THE SUFPORTINO CONNECTIVE TiSBUE CONTAINING CAPILLARIES 

Some are in intimate contact with the bands of connective tissue 
described above, but in most instances they are retracted from 
them. This is very likely due to fixation. Between the indi- 
vidual epitheUal cells there is no reticulum. Cell outline in 
many areas is rather indistinct, but for the most part the cells 
are polyhedral or rounded in shape. Most of the cells are ar- 
ranged in the form of sohd masses, but some of the latter are 
hollowed out and in a few sections apparent attempts at acinar 
arrangement are seen, the cells surrounding these imperfectly 



282 HARRY GOLDBLATT 

formed acini being of the cylindrical type. In many sections, 
what at first appear to be fairly well-formed acini are seen to 
contain in their lumen small masses of loosely arranged moder- 
ately nucleated connective tissue. These are formed apparently 
by the ensnaring of some of the connective tissue stroma by 
strands of epithelial cells which happen to be arranged in a 
circular manner. Many of these small areas of connective 
tissue within the pseudo-acini contain one or more small capil- 
laries. Wherever the connective tissue is small in amount the 
epithelial cells come into such intimate contact with the capil- 
laries that a spinous peritheliomatous appearance is the result. 
It is impossible to identify a well formed typical acinus, but in 
a few cases, owing to the cylindrical appearance of the cells and 
the pseudo-acinar arrangement one cannot rule out the possibility 
that this represents a feeble attempt at the formation of a tubule 
or gland follicle. 

Most of the tumor cells are polyhedral or rounded in shape, 
are approximately of the size of islet cells, in most instances 
definitely smaller than those of pancreatic acini, and show the 
same variation in size that the islet cells do. An occasional 
very large cell is seen but this is not uncommon in the normal 
islet. The cytoplasm of the cells is relatively moderate in 
amount, is definitely acidophilic, stains very lightly, and is 
granular. As in the nuclei of the islet cells, the chromatin is 
aggregated into a single large mass or several, usually three or 
four, small clumps. The examination of very many sections 
failed to reveal a single mitotic figure. 

The general architecture of the tiunor as described above 
bears a very striking resemblance to that of the islets of Langer- 
hans, so much so that if certain portions in which the masses 
and cords of epithelial cells are not large were seen in a normal 
pancreas they w^ould fulfill all the morphological requirements of 
an island. A careful study of the cells themselves confirms this 
very striking resemblance. 

SUMMARY 

That this tumor is of benign nature is deduced from the fol- 
lowing facts: It is definitely and completely encapsulated. 



ATYPICAL ADENOMA OF PANCREAS 283 

Neither the capsule nor the neighboring pancreatic tissue shows 
any signs of direct invasion by the new growth. There is no 
metastasis. There are no mitotic figures. The general arrange- 
ment of the tumor is not unlike that of some cellular adenomata 
found in other organs of the body. The epithelial cells of the 
tumor and their arrangement with relation to the vascularized 
trabeculae show unmistakable signs of diflFerentiation, since there 
is such a striking resemblance to the islets of Langerhans. In 
size, shape, staining characteristics, and the minute structure of 
the cytoplasm and nucleus the cells resemble very closely those 
of the islands. 

The tumor here described is in all respects like those reported 
by Nichols, Helmholz, Morse, and others. The only diflFerences 
are that this growth is much larger than any hitherto reported, 
and that by reason of its size and location it had exerted pressure 
upon the pancreatic ducts and had caused cUnical signs and 
symptoms suggestive of pancreatic insuffi.ciency. In this respect 
it differed from most adenomata hitherto described. Glycosuria 
was never found. 

CONCLUSION 

An adenoma of the head of the pancreas has been described 
which originated in an islet of Langerhans. 



(1 

(2 
(3 
(4 
(5 
(6 
(7 

(8 

(9 

(10 

(11 
(12 
(13 
(14 



REFERENCES 

Cecil, R. L. : Jour. Exper. M., 1911, xiii, 595. 

Heiberg, K. A.: Centralbl. f. allg. Path. u. path. Anat., 1911, xxii. 532. 
Helmholz, H. F. : Bull. Johns Hopkins Hosp., 1907, xviii, 185. 
Herxheimer, G. : Virchows Arch. f. path. Anat., 1906, clxxxiii, 228. 
HoRGAN, E. J. : Jour. Lab. and Clin. M., 1920, v, 429. 
Koch, K. : Virchows Arch. f. path. Anat., 1914, ccxvi, 25. 
LeComte, R. M. : Jour. Med. Res., 1913, xxix, 251. 
Morse, M. E. : Jour. Amer. Med. Assoc, 1908, li, 1075. 
MussEY, R. D. : Med. Clin. North America, 1919, iii, 681. 
Nichols, H. G. : Jour. Med. Res., 1902, viii, 385. 
Reitmann, K. : Ztschr. f. Heilk,. 1905, xxvi, Abt. f. path. Anat., 1. 
RoLLET, H.: Frankfurter Ztschr. f. Path., 1912, x, 268. 
Ssobolbw, L. W.: Virchows Arch. f. path. Anat., 1904, clxxvii, Suppl. 123. 
Weichselbaum, a.: Sitzungsbericht d. kaiserliche Akademie d. Wissen- 
schaft. in Wien, Mathem. natur. Klasse, 1910, cxix, 73. 



THE INFLUENCE UPON THE GROWTH OF TRANS- 
PLANTED FLEXNER-JOBLING RAT CARCINOMA OF 
HYDROGEN IONS AND OF VARIOUS SALTS IN DIF- 
FERENT CONCENTRATIONS 

KANEMATSU SUGIURA, HELEN MILLER NOYES, and K. GEORGE FALK 

From the Huntington Fund for Cancer Reaearchj Memorial Hoapttctl, and the 
Harriman Research Laboratory, The Roosevelt Hospital, New York 

Received for publication May 16, 1922 
1. INTRODUCTION 

In connectioii with the investigation of certain enzyme actions 
of extracts of malignant human and rat tumors, the results of 
which are published elsewhere (1), a study was made of the effects 
upon their subsequent growth of various treatments of the Flex- 
ner-Jobling rat carcinoma transplants before inoculation. The 
results presented here include the effects of solutions of different 
hydrogen ion concentrations and of some simple salts and their 
mixtures. 

2. EXPERIMENTAL METHODS AND RESULTS 

a. Tumor transplanUUion method 

Albino rats, both male and female, were used in this work. 
They were inoculated with the Flexner-Jobling rat carcinoma in 
the usual way, and the tumors were allowed to grow for a period 

of from six to seven weeks (2). The term ' for example, 

indicates that the tumor was derived from the 86th generation, 
series A. The rats were fed on wheat bread soaked in whole 
milk, fresh cabbage or carrots, and fresh tap water ad libitum. 
In carrying out the transplantations, six (or more) small 
particles of tumor tissue each weighing about 5 mgm., and one 
piece weighing about 50 mgm., were selected from the non- 
285 



286 K. SUGITJRA, H. M. NOTES AND K. O. FALK 

necrotic area of the rat tiunor and placed in 25-cc. portions of 
solutions of definite hydrogen ion concentration and salt con- 
tent previously sterilized in a steam autoclave at fifteen pounds 
pressure for fifteen minutes. The flasks containing the solutions 
and tumor fragments were allowed to remain in the ice box for 
definite intervals of time of from one-half to seventy-two hours. 
At the end of this period of time, the smaller pieces of tumor 
tissue were inoculated with a trochar into rats subcutaneously 
in the region of the right axilla. At the same time, the larger 
as well as some of the smaller pieces of tumor tissue were fixed 
in formaldehyde solution for histological examination. As a 
control, the same nmnber of animals were inoculated with un- 
treated tumor tissue immediately after its removal from the 
tumor-bearing animal. The progress of the transplanted tumors 
in the rats was recorded graphically by measuring them every 
week by means of calipers. 

6. Preparation of solriHons 

The pH values of the solutions were determined in most cases 
by means of indicators and the standard solutions recommended 
by Clark (3), and in some potentiometrically. The salts used 
were purified by crystallization in the usual manner whenever 
necessary. The compositions of the various solutions in which 
the tumor fragments were immersed are given in table 1. 

c. Growth of transplanted tumors after different treatments 

The complete results of the transplantation experiments are 
recorded in 39 charts. As it is manifestly impracticable, even 
if desirable, to reproduce this niunber of charts, the results will 
be given in the first instance in table 1 as far as possible. Al- 
though this method of presentation leaves much to be desired, 
especially in the recording of the detailed growths of the various 
transplants, it will serve to bring out the points which are to be 
emphasized. In order to present a partial idea, at any rate, of 
the various growths, a nmnber of more or less typical examples 
were chosen from the charts and are recorded in figure 1. 



FIGURE 1 




mim. 


9t f^MM-n^KW^ 


GronNth or tomor offer tncufoCHm. | 


-TfT 




Mtb. 


7«&« 


'••V 


ata^ 


J0 oto« 


45-«to« 


•M obv 


AO) 


II9€ 




^(0itn 




• 


• 


• 


• 


• 


• 


BW\ 


/3S6 


£3 


UAr 


34 


— 


— 


— 








BW 


laee 


$A 


buAt 


M 


— 


— 


• 


• 


• 


• 


BW 


laei 


6^ 


iufi, 


i4 


— 


— 


— 








CM 


IS62 


9.0 


iOSkr 


24 


• 


• 


• 


• 


• 




con 


ms 


1.0 


k£^ 


24 


• 


• 


• 


• 


• 




Ctfii 


1370 


6.0 


iuAt 


24 


• 


• 


• 


• 


• 




DM 


t4S5 


1.0 


NdCl 


24 


• 


• 


• 


• 


• 




OOii 


t4€2 


IJO 

t 


UCI 


24 


— 


— 


• 


'• 


• 




£M 


I5H 


7.0 


CaCk 


as 


• 


• 


• 


• 


• 




EM 


1550 


7.0 


CM 


5 


• 


• 


• 


• 


• 




FC»/) 


1551 


7.0 


CbOk 


5 


— 


— 


— 


— 







FM 


I560 


1.0 


CbOk 


10 


— 


— 


• 


• 


• 


• 


FW 


1562 


10 


CbCA 


10 


— 


— 


— 










OW 


1465 


iJO 


Q»C& 


24 


— 


— 


— 










HW 


ni6 


IX) 


Q/A 


24 


— 


— 


— 


• 


• 


• 


an 


1063 


10 


NccI 

nci 
each 


12 


• 


• 


• 


• 


• 


• 


JW 


xas 


TO 


i 

ack 


24 


— 


— 


— 


— 


— 




Km 


20H 


70 


NbCf 


12 


— 


— 


— 


— 


— 






















1 


H4+4- 



CA 



Fig. 1. Growth of transplanted Flexner-Jobling rat carcinoma after immer- 
sion in various salt solutions and in solutions of different hydrogen ion 
concentrations. 

* Concentration of salts as in Locke-Ringer solution. 

t Concentrations of salts three times those in Locke-Ringer solution. 

287 



288 



K. SUGIUBA, H. M. NOTES AND K. O. FALK 



Q 
w 

». 



't 

ft 






I 









g 



I 
2 



ft; 









'S i 

V a 

o 9 








week, 
ormal 
I after 
















in first 
two, n 
growth 
















a 

M 














rowth ] 
; other 
ne rat, 


< 






«vS g 








cA »i< O 


i 












O ^ ••» 


S 






i'S § 








•^ 08 « 




■s 

1 




•8-2 1 




5 




ats, 
here 
ton 
eek 




rB 


-•5S& 


5 


1 


5 


»4 ^ « ^ 




b 


t 


^ O B 


> 


M) 


^ 


§^5-2 






Gd 

o 


One ra 
tion, 


1 

o 
55 


3 


1 

o 


With fr 
norm 
grow 
secoE 


#H M ^ 


Is 


o 


{S 


o 


s 


o 


8 


O 14 ^ 


1- 










• 


c 




Time 

of 

immer- 

eion 


1 


s; 


S 


^ 




s 


s 


^^ 




rH 


o 


00 




00 


o 


\ 


•a 




»d 


• 


00* 




id 


• 






• 


• 


• 






• 






«» V 


c» 


o . 




o . 


o . 


1 

3 


a 




::8 


+ 29.6 
200 cc. 


+ 53.0 
200 cc 




lO ^ 


2§ 


g 


9 

1 




2| 


o5 


o'-. 




^^ 


o'5 


8 


•8 

! 




«-2 


Pl^' 


A4 p4 

2w^ 






Ph ^ 


8 




«o 


»o O 


i^q 




iSO 


4S.O 


§ 


1 


•i 


sw 


s« 


S^ 


•1 


sw 


S« 


s 


'8 




§5 






1 






s 




g 


^s 


JJS 


^s 


g 


^s 


^s 


i 






s 


8 


s 




3 


s 


TUUOB 
U8BD 


t 


CQ Q 

8 g 




i 


S 


i 


PQ 

s 


1 








< 


S 8 ^ a S 


to 


IQ 


lO 


»o 


U9 


lO 


»o 


S 






*"-*» 


y-s 


^-s 


^^ 


/— s 


s 


fM 


s 


CO, 


^ 


s 


5S 


fc. 






^ 


< 




< 




< 




n 


PQ 




p:) 





GROWTH OF FLEXNER-JOBUNO HAT CARCINOMA 



289 



I 

s 

OB 

1 









O 






•»'« 


«« o 
M d 


« 


o 


o - 


«> « 


> s 


^ o 
^ o 


••9 th 


•»> h 


g5 


1^ 



« . 
o o 

S g 

iS.O 

s 






<s 



I 



3, 



R - 

p U 9} 

©•a § 



fe - 

O h o 

o pd ^ 

Sab 



I 

s 

S 
^^ 

S 

*•! 
IS. 









00 

(d 

.g 

no 

o C 



53 



i 


s 


8 


s 


s 


s 


S 


s 


s 




;; 


s; 


s 




S 


;s 


04 

00 




e 

■ 


o 

• 


• 

CD 




• 


o 

• 









o 



00 
73 






M a to 



8 



n 

1= 



o o 

Sg 

£'f 

»p o 



g 



d 

V 

o 
u 



to 






CO 



o 
o 

hi 

s 










290 



K. SUarDBA, H. M. NOTES AND K. O. FALE 



a 



I 

o 
55 



1 


1 


o 


o 


b) 


Si 


s 


:2 


9* 


9* 


(S 


iS 



5 

o 



I 

o 
a 

o 

• ^ 

I 



o 



iS s 



! 



o 

2 



r 






g 



o 2 

s i 



I 




i 



8 8 S S S 8 § 



s 



3 
9 



8 



s 

A 

i 

H 
O 

p 

5 
§ 



9 



9% I 



« a 

o 

a 



§! 



lO 



H 



a 

I 

•s 

g 






o 
g 



o 

OS 

o 
i 



o 

08 

o 



o 
g 






*o 



■d 



o 



s 



o 

OS 
O 



i 



s 



Hi 



Q 



I 



I 



Q 



Q 






P3 

3 






3 



I 



3 






lO 



ao 



lO 



•O 



lO 



lO 



»o 



a 

a 

■ 






GO 



OK 



^ 






1^ 



GROWTH OF FLEXNEB-<rOBLIN6 RAT CARCINOMA 



291 




'^M 



o 





^ 






S 




>% 


M 










3 


e 




Q* 


« 




^ 


,jd 




u 


•liS 




•*a 


o 




OB 






S 


> 




■ •« 


o 




>f 


&) 




1 

r-4 


o 




«0 


d 




t 


•o 




49 


•S 




Ii4 


•d 




o 


o *? 


5 


1 


S 3 


o 

bO 

O 


O 


o 


^ 



I 

o 



S 



S 



S S 



;$ 


s 


s 


12 


8 


8 


o 

• 


o 

• 


o 

• 

1^ 


o 

m 


O 

• 


O 

• 



o 

OS 
O 



i 



J9 

o 

o 
O 



49 

o 

§ 



O 

OS 

O 

i 



J9 
o 

g 






'Sp's 





Sz; 



58 



I 






o 



E ^ E 



< 
S 



o 






O 






1 



S 



g 

f^ 



S 



I 



PQ 
S 



I 



s 



>o 



»o 



00 



to 



»o 



lO 






o 



O 



n 









Q, 



^. 



292 



K. SUOITTRA, H. M. NOTES AND K. O. FALE 



9 



O 

d 
"2 

§ 



s 







o 
55 






<§ 



o 



o 

2 



•8 

I 

s 



s ^ 



*§ 



<s 



5 
g 

O 



o 

Sz: 




I 



& 



s 



s 






ls 



ss 



8 e 



n 

A 



a 

•I 

8 

a 






o 

u 

g 



OS W eS 






sss s 



o 
o 






OS 

O 



8S 



S 



o 



Q 



gs g 



8 






§ 



PQ 

8 



Q 



8 



o 



2P 



w 



»o 



to 



a 



s s § 



»-9 












GBOWTH OF FLEXNER-JOBLINQ RAT CABCINOMA 293 

The data given in table 1 are perhaps sufficiently self-explana- 
tory. The experiments were allowed to proceed six to seven 
weeks in every case. In the column showing percentage of 
positive inoculations, whenever the growth was delayed for a 
period greater than one week, it was calculated as a positive 
inoculation, but the number indicating the percentage was en- 
closed in parentheses. Figure 1 shows the relative growths under 
the different conditions more clearly. 

The results of the various treatments on the growth of the 
transplants can be summarized in a comparatively brief manner 
as follows: 

1. After immersion in potassium phosphate buffer mixtures of 
various hydrogen ion concentration for twenty-four hours, the 
following results were obtained : No effect was observable at pH 
7.0, the growths being practically the same as in the control, "no 
growths at all were obtained at pH 5.8 and 5.1 ; at pH 8.2 partial 
inhibition and delayed growths were found; and no growths at 
pH 8.8. 

2. Varying the concentration of the buffer mixture (from 
approximately 0.6 per cent to 1.1 per cent) at pH 7.0 (series 
CIO, Cll, and C12) and immersing for twenty-four hours did 
not produce any effect, the growths being normal in all the cases. 

3. Sodium chloride (0.15 M or 0.9 per cent) at pH 7.0 had 
no effect in twenty-four hours on the growth of the transplants; 
seventy-two hours immersion resulted in complete inhibition of 
the growths; lithium chloride at pH 7.0 in twenty-four hours 
caused partial inhibition. 

4. Calcium chloride at pH 7*0 inhibited or retarded the growths 
of the transplants. Immersion for one-half hour had no effect; 
for five and ten hours, partial inhibition or retardation; for 
twenty-four hours and longer, complete inhibition. 

5. Immersion in a Locke-Binger solution for seventy-two 
hours did not affect the growths. Immersion in a solution 
containing three times the concentration of the sodium, calcium, 
and potassiiun chlorides in the Locke-Ringer solution for twenty- 
four or seventy-two hours resulted in complete inhibition of the 
tiunor transplants. 

TBV J017BNAL OP CANCSB BX8KAB0B, VOL. VI, WO. 4 



294 K. BUGITTBA, H. M. NOTES AND K. G. FALK 

d. Histological examinations of tumors 

The histological examinations of a number of the tumor trans- 
plants after immersion in the various solutions were made by 
Dr. James Ewing, to whom the writers wish to express their 
thanks. The general structure after inmiersion in the phosphate 
mixtures of various hydrogen ion concentrations showed slight 
hydropic degeneration in the case of pH 7.0, but with the more 
acid and more alkaline solutions the tumor cells had undergone 
marked hydropic degeneration, being larger and considerably 
swollen. The microscopical structure of this timior is shown 
in plates 1 and 2; in plate 1 an untreated section, and in plate 2 
one treated with a solution of pH 5.8. The histological com- 
parison of untreated tumor and tissue which had been immersed 
in solutions of sodium chloride, lithium chloride, and calciimi 
chloride also showed some interesting facts. In those treated 
with sodiimi or lithium chloride, there was considerable accumu- 
lation of fluid between the tumor cells and many nuclei appeared 
slightly shrunken. On the other hand, tissue which had been 
treated with calcium chloride showed imbibition of fluid in the 
intercellular spaces and the cells themselves were shrunken. 
The nuclei stained more densely and the chromatin was condensed 
to a central mass within the nuclei. The C3rtoplasm showed 
hydropic vacuolization (plate 3). A microscopical examination 
of tumor tissues was made after immersion for seventy-two hours 
in a Locke-Ringer solution, and also of the same timior tissue 
treated by a modified Locke-Ringer solution in which the salt 
content was increased three times. Under both conditions the 
tissues showed marked hydropic degeneration, but the proliferat- 
ing capacity of the tissue treated with normal Locke-Ringer 
solution was not altered; while under the second condition 
proliferation was completely inhibited. 

3. DISCUSSION 

Li comparing the results of the tumor transplanting experi- 
ments, it is necessary to consider not only the contents of the 
solutions in which the tumor fragments were immersed before 



GBOWTH OF FLEXKEB-JOBLING BAT CABCINOMA 295 

the inoculations, but also the lengths of time of the immersions. 
The temperatures were kept fairly constant, otherwise these 
also would have to be included. It will be seen that in discussing 
the contents of the solution, not only must the salt and its 
concentration be considered, but that the hydrogen ion concen- 
tration is of the greatest significance* 

The favorable effect of immersion for twenty-four hours in 
solutions of pH 7.0 as compared with the more acid or more alka- 
line solutions is striking. Immersion in the pH 8.0 solution 
appeared to be less harmful than immersion in the pH 6.0 
solution. Complete inhibition of growth was caused by the 
pH 6.0 and pH 9.0 solutions in twenty-four hours. From a 
histological examination of the tumor cells after immersion in 
solutions more acid or more alkaline than pH 7.0, it was found 
that there was a hydropic degeneration of the cells similar 
to that of tumor cells resulting from exposure to radiimi or to 
aj-ra3rs. 

The action of the salts on the timior tissue before inoculation 
is of interest. Sodiimi chloride had no effect in twenty-four 
hours, but exerted a harmful action in seventy-two hours. lith- 
ium chloride produced a small inhibiting action in the twenty- 
four hoxu: treatment. Calciiun chloride caused a very strong 
iohibiting or retarding action on the subsequent growths even 
in ten hours. The use of potassium salts in the buffer mixtures 
showed that potassium exerted no specific retarding influence. 

Cramer (4) several years ago studied the effects of sodium 
chloride and calciiun chloride solutions on the growth of a trans- 
plantable mouse carcinoma. He allowed isosmotic solutions of 
these salts to act on the cells for one to two hours before inocula- 
tion. Marked inhibition of growth was shown by the cells 
immersed in the calciiun chloride solution, practically none by 
those immersed in the sodium chloride solution. He found that 
the harmful effects of the calcium chloride could be overcome 
by subsequent immersion in sodium chloride solution, and that 
the action of the calcium chloride was only transient. 

Although the experiments of Cramer are of interest in con- 
nection with the present work in showing a parallelism between 



296 K. BUGIUBA, H. M. NOTES AND K. G. FALK 

the harmful action of the calcium salt and the less harmful action 
of the sodium salt, the following differences in the manipulations, 
aside from the use of mice in the one case and rats in the other, 
may be pointed out. In the present work, the hydrogen ion 
concentrations of the solutions in which the tumor fragments 
were immersed were controlled; in Cramer's work they do not 
appear to have been controlled. In the former, the inoculated 
animals were observed for six to seven weeks after inoculation; 
in the latter for two weeks. In the former, also, the times of 
immersion were varied for a number of the treatments, and cer- 
tain salt mixtures were used. In spite of these differences, the 
results of Cramer offer valuable contributory evidence in connec- 
tion with the transplantation phenomena. 

The results of the transplantation experiments described in 
this paper are evidently due to cell destruction of some form. 
In view of the harmful action of the salts separately, and the 
harmlessness of the ''balanced" mixture, the action on the cell 
membrane or wall appears to be the dominating phenomenon. 
Destruction of the membrane or cell waU, or perhaps better, 
modification of its permeability, destroys the reproductive 
power of the cell. The actions of small amounts of acids and 
of bases, and of various salts and mixtures of these, have been 
described in connection with other cell structures a nimaber of 
times. In the present work, the effects of solutions of different 
hydrogen ion concentrations have been made more definite, and 
the relative action of sodium and calcium salts at a definite hy- 
drogen ion concentration brought out clearly. The explanation 
of these actions is, however, the same as that developed in other 
connections by a nimiber of workers. 

The work on the protease actions of malignant human and 
rat timior extracts presented elsewhere (1) shows some interest- 
ing similarities and differences as compared with the transplanta- 
tion phenomena presented here. While it is true that the causes 
underlying the two sets of phenomena are probably of entirely 
different nature, the protease results will be outlined briefly in 
comparison with the transplantation results. 



GBOWTH OF FLEXNEB-JOBLING RAT CARCINOMA 297 

The optimum hydrogen ion concentration for protease action 
and the favorable medium in which tumor fragments on immer- 
sion retain their ability to grow when transplanted, correspond 
very closely to that represented by pH 7.0, unfavorable condi- 
tions for both being reached more rapidly on the acid side than 
on the alkaline side. Chlorides of the alkalies in fairly dilute 
solution and within short time limits, did not affect appreciably 
the protease action or the ability of tumors to grow when trans- 
planted, but calcium salts under similar conditions exerted a 
very marked inhibiting or retarding action on both. On the 
other hand, a definite difference is observed with a mixture of 
sodium chloride, calcium chloride, and potassium chloride in 
the concentration found in Locke-Ringer's solution. Retarda- 
tion of the protease action was caused to the same extent that 
the salts separately would cause such action, while with the 
transplantations, the salts exerted an antagonistic action toward 
each other in the sense that no inhibition of the tumor growth 
was observable in the subsequent inoculations. 

The two sets of actions, therefore, can be ascribed to different 
causes; the transplantation results in the first instance to salt 
actions on the permeability of the cell membranes, the enzyme 
results to chemical actions of unknown nature on the enzyme 
molecule or enzyme grouping of some molecule. The compara- 
tive actions of the sodium and calcium salts parallel each other 
when used alone, but not when used in mixtures in certain pro- 
portions and concentrations. The parallelism, even if due to 
different causes, as is probable, is of interest. 

Certain facts may be mentioned in this connection, even if 
they have no direct connection with the phenomena described 
in this paper. In reviewing past work on the inorganic constit- 
uents of neoplastic tissue, it may be stated that the calcium 
content of rapidly growing tumors was found to be small. Cal- 
culating the potassium-calcium ratio of such tumors, large 
concentrations of potassium as compared with the calcium were 
found, while in old, necrotic, or slowly growing tumors, the con- 
centration of calcium was considerably greater in comparison 
with that of the potassium (5). 



298 K. SUGIURA, H. M. NOTES AND K. G. FALK 

Finally, it may be stated that while a definite connection 
may exist between the conditions affecting the permeability 
of cell membranes, the factors which influence the activities of 
intracellular enzymes, and the relations of the inorganic con- 
stituents of neoplastic or other tissue, such a connection is still 
obscure and can only be hinted at until more definite experi- 
mental evidence is available. 

4. SUMMABY 

The growth of the Flexner-Jobling rat carcinoma was investi- 
gated after grafts had been unmersed in solutions of various 
salts of different hydrogen ion concentration. Calcium strongly 
inhibits growth, and a pH of 6.0 appears to be more harmful 
than pH 8.0. 

The growth of the tumors was compared with the protease 
action of extracts of malignant human and rat tinnors, and simi- 
larities and differences in these phenomena indicated, as well as 
possible explanations for them. 

REFERENCES 

(1) J. Biol. Chem., 1922, liii, 75. 

(2) SuGiURA, K., AND BENEDICT, S. R. 1 Jour. Canccr Res., 1920, v, 373. 

(3) Clark, W. M. : The Determination of Hydrogen Ions, 1920. 

(4) Cramer, W.: Biochem. Jour. 1918, xii, 210. 

(5) Beebb, S. p.: Am. Jour. Physiol., 1904, xii, 167. 

Clowes, G. H. A., and Frisbie, W. S. : Amer. Jour. Physiol., 1905, xiv, 173. 
Waterman, N. : Arch. Neerlnd, Physiol., 1921, v, 305. 



PLATE 1 

The plates show three photomicrographs taken with the same magnification. 
Plate 1 shows untreated tumor tissue of the Flexner^obling rat carcinoma. 



GROWTH OF FLEXNER-JOBLINO RAT CARCINOMA 



PLATE 2 

Plate 2 represents the same tumor after immersion for twenty-four hours in 
the phosphate mixture, having the pH 5.8. 



300 



OROWTH OF FXEXNEH^OBLINO RAT CARCINOMA 



PLATE 3 

Plate 3 shows the results after immersion of the same tumor for twenty-four 
hours in the 0.078 M calcium chloride solution, having the pH 7.0. Microscopical 
examination of the treated tissues showed considerable degenerative changes. 
(For further reference see text.) 



302 



OROWTH OP FLEXNER^OBLIN'O RAT CARCINOMA 



PRIMARY SPONTANEOUS TUMORS IN THE KIDNEY 

AND ADRENAL OF MICE 

STUDIES ON THE INCIDENCE AND INHERITABILITY OP 

SPONTANEOUS TUMORS IN MICE 

Seventeenth Communication 

MAUD SLYE, HARRIET F. HOLMES, and H. QIDEON WELLS 

From iht Otho 8. A, Sprague Memorial Institute and the Department of Pathology 

of the University of Chicago 

Received for publication May 8, 1922 

Primary tumors of the kidney occur not infrequently through- 
out the animal kingdom, and, in general, seem to exhibit the 
variations and peculiarities seen in human renal tumors. As an 
indication of the comparative pathology of renal tumors the 
following review of the literature is presented: 

Mouse: Few cases of primary renal tumors have been de- 
scribed among the great numbers of other tumors foimd in this 
species* Tyzzer (1) found 4 renal growths, which he interprets 
as hypernephromas, among 83 primary tumors in mice. One 
of these mice, an old female, had also an adenoma of the limg, 
a carcinoma of the ovary, and a lymphoma infiltrating both 
kidneys; another had an adenoma in the lung. At the time 
this paper was written the interpretation of hypernephroma was 
more liberal than at present, and the illustrations might now be 
interpreted by many pathologists as papillary adenoma of the 
kidney, which was Tyzzer's original diagnosis in two of his cases. 

Among the 300 mice with spontaneous tumors described by 
Haaland (2), there were but two renal tumors. One was a large 
growth in the kidney of a twenty-two months old male mouse, 
and had invaded a vein but produced no metastases; micro- 
scopically the structure was that of an adenocarcinoma. The 
second tumor, in a mouse of the same sex and age, resembled 

305 



306 MAX7D SLYE, H. F. HOLBCBS AND H. G. WELLS 

perfectly in structure a hiunan hypernephroma; there were no 
metastases> although the lung contained an adenoma. Inocula- 
tion of the hypernephroma into 40 mice was without result. 

Miuray (3) also described a i^indle-cell growth surrounding 
the kidney without infiltrating it, apparently a primary retro- 
peritoneal sarcoma and related to the kidney by position only. 

Further than the above we can find no reports of primary renal 
tumors in mice, although Stumpf (4) has contributed a discussion 
of the behavior of carcinomas inoculated into the kidney. 

Rats: Of 103 tumors found in 100,000 rats autopsied in plague 
work by McCoy (5), 11 were in the kidney, and were classified as 
4 adenomas, 6 carcinomas*, and 1 papill(»na. Woolley and Wherry 
(6) in 23,000 rats found 22 tumors, of which 3 were in the kidney; 
all were of renal-cell type. Bullock and Rohdenburg (7), in a 
compilation of the literature on rat tumors, found 123 (including 
those cited above) of which 7 were fibroadenomas^ 1 a papilloma, 
and 8 carcinomas of the kidney. Among 32 cases of their own 
in laboratory white rats^ there were 10 adenomas of the kidney. 
Therefore, of 155 rat tumors 26 were of renal origin. A further 
case of carcinoma of the kidney in a wild rat was reported by 
Beatti (8) . Loewenstein (9) described tumor-like papillary out- 
growths arising in the pelvis of the kidney, as well as in the 
ureter and bladder, of rats infected with Trichodes crassicauda 
specifica. 

Since sarcomas are much more common than carcinomas in 
rats, the fact that all the renal tiunors in rats are of epithelial 
structure is of interest. 

Squirrels: In 250,000 ground squirrels (Citellus beecheyi) au- 
topsied in plague work, McCoy (10) found 8 tiunors, of which 
one was described as an angiosarcoma of the kidney. A case of 
hypernephroma in the kidney of a grey squirrel (Sciurus caro- 
linensis pennsylvanicus) was reported by Fox (11). 

Birds: Fowls are rarely subject to renal tumors. Of 880 
examined by Curtis (12), 79 had tumors^ 5 of which were in the 
kidneys; but these were not microscopically corroborated and 
there is reason to doubt their true neoplastic nature. In 34 
cases of tumor in fowls compiled by Wernicke (13), not one was 



TUMOBS IN KIDNEY AND ADRENAL OF MICE 307 

primary in the kidney although in 3 cases renal metastases were 
found. In 862 autopsied fowls Btirget (14) found 12 tumors of 
which none was in the kidney, although a sarcoma of the ovary 
had produced metastases in the kidney in one case. 

The review of the literature on tumors in fowls and birds by 
JoQst and Emesti (15) who collected 112 cases and added about 
50 more, reports no further renal tumors. 

Other birds seem to have renal tumors perhaps more frequently, 
as is indicated by the following reports: White (16) described a 
fibrosarcoma in the kidney of a goose. Fox (17) reported in the 
kidn^ of a male chestnut eared finch (Amadina castanotis) a med- 
ullary carcinoma, becoming scirrhous in places^ with metastases 
to the lungBy and three oases of benign adenoma of the kidneys in 
undulated grass parrakeets (Melopsittacus imdulatus). He had 
previously reported (18) two cases of papillary cystadenoma and 
also a ''cyst adeno-carcinoma papilliferum" (19) ii^ the same 
species; besides an adenocarcinoma of the kidney in a saffron 
finch (Sycalis flaveola), (20) and a spindle-cell sarcoma of the 
left kidney with metastasis into the left tibia in a scaly ground 
dove (Scardapella squamosa) (21). Seligmann (22) reported 
as a diffuse carcinoma a growth which involved both kidneys, 
with metastases in the liver and mesentery, in a Chilian pintail 
(Dafila spinicauda) which was .twenty-sbc years old. Baird (23) 
reported a case of keratinizing epithdioma in the kidney of a 
fowl, and Borrel and Masson (24) also described a renal tiunor 
in a fowl which showed both cylindrical and squamous elements^ 

Domesticated mammals. Rabbits, which are less subject to 
tumor than most mammals, seem to have a relatively large 
proportion in the kidney, for Scott (25) stated that of 39 new 
growths reported, 5 were benign renal adenomas, resembling 
Wilms' tumors in structure. 

Swine also seem to be particularly subject to renal tumors, for 
of the 12 cases of tumors in swine collected by Sticker, 7 arose in 
the kidnqr, and other oases reported since then indicate the same 
tendency. Many of these tumors were of the nuxed embryonal 
tumor type. 



808 MAUD SLYE, H. F. HOLMES AND H. G. WELL8 

In horses renal tumors are far from rare. In Sticker's (26) 
compilation, of 509 cases of malignant tumors in horses 37 were 
in the kidneys. McFadyean (27), in 63 cases of tumor in ani- 
mals included 5 in the horse kidney, none with metastases. Of 
142 equine neoplasms among 77,224 horses slaughtered in Japan, 
there were 9 in the kidney, as compared with 49 in the testicle 
(Kimura) (28). 

According to Trotter's (29) figures, renal tumors are much 
less frequent in cattle, for of 305 cases but 1, a colloid cancer, 
arose in the kidney; but Sticker's figures show 10 of 78 bovine 
timoLors in the kidnejrs. Stdinke (30) described as hypernephroma 
a tumor of the kidney in a cow. Roussy and Wolf (31) in their 
review on cancer in animals, gave a picture of a tumor from a 
bovine kidney resembling a hypernephroma, and stated that 
cancer of the kidney in horses resembles that in man, while in 
swine the adenosarcoma of the kidney is among the most common 
tumors. Cadiot (32) quoted a case of an enormous cancer of the 
kidney in a mare. 

No reports of cases of renal tumors in sheep can be found. 

Of 766 primary cancers in dogs in Sticker's tables 19 were in 
the kidney. McFadyean also described two renal tumors in 
dogs^ each case exhibiting metastasis. No renal tumors appeared 
in the 21 cats with tumors in Sticker's lists, nor in Roffo's 7 cases 
(33) nor in Murray's 11 cases (34) but Teutschlaender mentioned 
the cat (Eater) on his list of animals in which carcinoma of the 
kidney has been reported. Murray also described 48 cases of 
tumors in dogs, 12 in horses, and 18 in cows> with none arising in 
the kidneys. 

We are indebted to Dr. L. E. Day for his summary of 316 
tiunors found among 2000 animal q>ecimens sent to the Chicago 
Laboratory of the Bureau of Animal Industry. These do not 
represent all the timiors that are observed in the slaughter house, 
but merely specimens sent to the laboratory by the injectors 
when in doubt concerning the diagnosis. Among 175 tumors of 
cattle there were no renal tumors although there were four adrenal 
tiunors, two diagnosed as sarcomas and two as hypernephromas. 
Among 93 tumors in pigs no less than 52 were in the kidney. 



TUMORS IN KIDNEY AND ADRENAL OF MICE 309 

47 being diagnosed as embryonal adenosarcomas and 5 as sar- 
comas ; there were no adrenal tumors. In considering the tumors 
in swine it is to remembered that nearly all swine are 
slaughtered before they are two years old, so that not many of 
the slaughtered animals have reached an age for developing 
carcinomas. Among 48 tumors from sheep none whatever Wfere 
found in the kidney or adrenal. These figures emphasize the 
infrequency of renal timaors in cattle and sheep and the fre- 
quency of mixed tumors of the kidney in swine, which resemble 
the typical mixed tumors of the human kidney. 

Other incidental cases are: Sarcoma of the kidney in a swine 
(Hamburger) (35), the report of which leaves some uncertainty 
whether the growth may not have been leukemic. 

Papillary adenoma in the left kidney in a mongoos lemur 
(White) (36). A tumor involving the kidneys, liver, lungs, and 
ovaries of a python, the primary site not bemg determined 
(Bland-Sutton) (37). A leiomyoma in the kidney of a horse, 
described by Wells (38). Pick (39) described a soft adenoma 
involving both kidneys of an eel. Williams (40) stated "A 
malignant renal tumor (derived from an adrenal 'rest') has been 
described by Bland-Sutton in a marmot." In Teutschlaender's 
compilation (44) were listed the following cases from the literature : 
Hemangioma of the kidney in a horse; carcinoma of the kidney 
in fowl, frog, pike, and buzzard. 

While the above compilation is certainly not entirely complete, 
it serves to bring together much of the literature on the com- 
parative pathology of renal tumors, and to indicate their general 
distribution and frequency. 

Adrenal tumors. We can find no reports of primary adrenal 
tumors in mice, despite the great numbers of mice that have been 
examined at autopsy and the not infrequent occurrence of ad- 
renal tujnors in other species. Thus, Kimura (42) found in the 
Uterature records of adrenal tumors in 24 horses and 46 cattle, 
and he himself found 5 adrenal tumors among 46. tumors from 
horses. The extensive discussion of the comparative pathology 
of adrenal tumors by Steinke (43) mentioned tumors only in 
horses and cows, commenting on the absence of such tumors in 

THE JOCRNAIi OP CAXCER BBSEARCB, VOL. VI, NO. 4 



310 MAUD SLYE, H. F. HOLMES AND H. G. WELLS 

other species except for an adrenal tumor of parasympathetic-cell 
character in a two-year-old swine, reported by Klawitter. 

There is, however, the interesting case reported by Smallwood 
(44) as a carcinoma of the kidney of a frog and believed to be 
derived from the adrenal tissues. Murray, who reexamined this 
specimen, corroborated its carcinomatous nature, but hi^ state- 
ment suggests that he was not convinced that it developed from 
adrenal rather than renal cells. The illustrations in Smallwood's 
article suggest a papillary renal tumor. 

Fox (45) described as a hypernephroma a tujnor that arose in 
the adrenal of a California hair seal (Zalophus calif ornianus) , 
and also described a hypernephroma in the adrenal of a brown 
cebus (Cebus fatuellus) (46). 

As stated previously, in the 316 animal tumors examined by 
Day in the Chicago Stock Yards, but 4 arose in the adrenal, all 
in cattle, 2 being diagnosed as sarcoma and 2 as adenoma. 

RENAL TUMORS IN MICE IN THE SLYE STOCK 

In 33,000 autopsies performed on mice of the Slye stock, but 
16 cases of unquestionable primary tumor arising in the kidney 
have been observed, supporting the impression that the kidney of 
the mouse is not among the common sites of primary neoplasm. 
These renal tumors are, according to the histological evidence, to 
be classified as follows: 1 carcinoma, 3 adenomas, 1 hyperne- 
phroma, 7 sarcomas, 3 mesotheliomas, and 1 sarcoma in the 
renal pelvis. It will be noted that we have observed no in- 
stances of the mixed tumors of the kidney that are so common 
in man and some other species, nor have we had any papillary 
tumors of the renal pelvis. The features presented by these 
several types of renal tumor are best described by giving briefly 
the findings in each of the several cases that we have observed. 

EPITHELIAL TUMORS OF THE KIDNEY 

1. Primary carcinoma of the kidney (fig. 1). Female mouse 
(1934) . This mouse showed a white solid mass in the right kidney 
about the size of a normal kidney. Beyond this there were no 



TUMORS IN KIDNEY AND ADRENAL OF MICE 311 

abnormalities found, except that the spleen was about three times 
the normal size. The left kidney was normal, and no metastases 
were found. Microscopically the tumor is composed of cells of 
epithelial type, but without the characteristics of either adrenal 
cells or of tubular epithelium of the kidney. They are arranged 



Fia. 1. Primart Carcinoua of the Kidnet 
The junction of the compressed renal tisaue and the tumor is shown. Appar- 
ently this tumor is derived from renal epithelium. Mouse 1934. X 110. 

in large, pseudo-alveoli with much central necrosis, so that the 
persistence of the tumor cells near the bands of stroma gives in 
places the impression of a papilloma. Although the growth 
does not in&ltrate the kidney very much, it has no capsule of its 
own, infiltrates the renal capsule in places, and is undoubtedly 



312 MAUD SLYE, H. F. HOLMES AND H. G. WELLS 

malignant. Evidently it is a tumor derived from the renal 
epithelium and may be properly designated as a carcinoma, 
although the term mesothelioma might also be appropriately 
used. Presumably it is in an early stage of malignancy in view 
of the relatively small amount of extension of the growth. 

2. SoUd adenoma arising in bilateral cystic kidneys. Male 
mouse (9907) , age two years, thirteen days. Both kidneys were 
converted into masses of small cysts of various sizes, resembling 
the congenital cystic kidneys, although no cysts were found in 
the liver. Some of the cysts contain colloid masses, and between 
them are occasional foci of small round cells. There remains 
more kidney tissue in a functional condition than is usually 
seen in fatal himiail cases of congenital cystic kidney, but this 
tissue is far from normal, the tubules containing many hyaline 
casts, the interstitial tissue being infiltrated with round cells, 
and many of the glomeruU being more or less hyalinized. In the 
lower pole of each kidney was a fleshy nodule about 3 to 4 mm. 
in diameter. Microscopically these nodules are composed of a 
solid tissue, made up of groups of large epithelial cells somewhat 
resembling renal epithelium. About these is a delicate stroma 
containing some collections of small round cells. The nodule is 
distinctly encapsulated and seems to be in the renal substance 
rather than within one of the cysts. In many respects the 
structure resembles that of the benign ovarian adenomas of 
mice (47). No abnormalities of importance were noted in the 
other organs. 

3. Adenoma of the kidney (fig. 2). Male mouse (10220), which 
died with advanced sarcosporidiosis ; had marked chronic nephri- 
tis, both kidneys being large and nodular, with atrophic areas 
alternating with areas of swollen and dilated tubules, some of 
the latter being distinctly cystic. Beneath the capsule of one 
kidney was a nodule 4 nun. in diameter, with a well defined 
capsule. It consists of a mass of large epithehal cells with soUd 
cytoplasm, arranged in large cords with a very delicate stroma 
between them. The arrangement of the cells resembles that of 
the adrenal adenomas, and a diagnosis of benign hypernephroma 
might readily be made. The cells are less vacuolated than those 



TUMORS IN KIDNEY AND ADRENAL OF MICE 313 

of the typical adrena] cortex, having a denser cytoplasm, and are 
not dissimilar to the epithelial cells found lining the distended 
renal tubules; it seems probable, therefore, that this adenoma is 
derived from renal epithelium. There are some clefts containing 



Fio. 2. Adenoma op Kidney 
This somewhat resembles a tumor derived from adrenal cells, but is believed 
to btkve been derived from renal epithelium. Mouse 10220. X 65. 

colloid material, and numerous calcific granules in the form of 
calcospherites. 

4. Adenoma of the kidney. Male mouse (24073), with a 
large liver cyst containing a tapeworm, exhibiting in one kidney 
a solid subcapsular nodule 5 mm. in diameter, with a dehcate but 
definite capsule. This is composed of a solid mass of lai^e 



314 MAUD SLYE, H. F. HOLMES AND H. G. WELLS 

epithelial cells with abundant foamy cytoplasm and small dark 
nuclei. The cells are in large cords or alveoli with a very deli- 
cate stroma between them. There are numerous small foci of 
calcification of the tumor tissue. The foamy character of the 
tumor cells, tnrcther with their arrangement, suggests an adrenal 



Fig. 3, Renal Hypernephroma 
This growth entirely replaced one kidney, but produced do metastases; it 
corresponds in structure to human hypernephromas. Mouse 3639. X 110. 

origin for this growth, but the post-mortem changes have so 
altered the nuclear and cytoplasmic details that this cannot be 
determined positively. In places there is a tendency to tubular 
arrangement of the tumor cells, suggesting a renal origin. Out- 
side the tumor the kidney shows an advanced chronic nephritis 
with many hyaline and fibroid glomeruli. 



TUMORS IN KIDNEY AND ADRENAL OF MICE 315 

5. Hypernephroma in the kidney (fig. 3). Male mouse (3639) 
with no other lesions of significance. The left kidney was re- 
placed by an encapsulated, hemorrhagic mass, measuring 22 x 22 
X 18 mm. The right kidney was slightly enlarged and soft, 
and showed a moderate degree of chronic nephritis. No remains 
of renal tissue are found in the tumor mass, which presents the 
typical appearance of a hypernephroma. More than half the 
mass contains no cells, being composed of the residue of old 
hemorrhages and necrosis. The living portions consist of large 
foamy cells arranged in cords and alveoli with a delicate stroma. 
It is completely surrounded by a capsule which shows no inva- 
sion by tumor cells. In all respects this growth corresponds 
perfectly to the human renal hypernephromas. No metastases 
can be found. 

In the hilum of the right kidney, attached to a large artery, is 
a nodule about 2 mm. in diameter which, in the center, resembles 
a small leiomyoma, but about it is a mass of granulation tissue. 
It bears no resemblance to the hypernephroma and its nature is 
unknown. 

In view of the fact that chronic nephritis is one of the com- 
monest diseases in mice, and appears in forms often quite similar 
to nephritis in man, it is strange that we have found so few in- 
stances of epithelial neoplasms in mice, especially the benign 
adenomas which are so often found in human kidneys showing 
chronic nephritis. 

SARCOMA OF THE KIDNEY 

The diagnosis of sarcoma, always difficult and often unsafe, 
is particularly dubious in the case of the kidney which presents 
so many non-sarcomatous growths that resemble sarcoma, and 
especially in the face of the statement made by Ewing (48): 
"Birch-Hirschfeld's group of adeno-angiosarcoma, derived from 
Wolffian remnants, and the lipomyosarcomas remain, however, 
the only well-defined varieties of renal sarcoma which have been 
fully divorced from a probable epithelial origin.'' Nevertheless, 
since the kidney contains connective tissue it is perfectly possible 
for sarcomas to arise therein, and in this mouse material we have 



316 MAUD SLYE, H. F. HOLMES AND H. G. WELLS 

several specimens for which only the diagnosis of sarcoma can 
be made, after excluding all other possibilities. In doing this 
we have carefully eliminated numerous growths of doubtful 
character, some of which also may really be sarcomas.^ These 
cases are briefly described as follows: 

6. Bilateral sarcoma of kidneys. Male mouse (7667), with 
both kidneys symmetrically enlarged to equal size, about 17 x 10 
X 9 mm. They contained little recognizable kidney tissue, and 
were for the most part infiltrated by a fleshy, pinkish white 
tissue which involved equally the cortical and medullary portions. 
No lesions were found elsewhere. Microscopically both kidneys 
show infiltration replacing about 80 to 90 per cent of the renal 
elements. The neoplasm is composed ever5n5vhere of slightly 
oval cells, a little larger than lymphocytes, with deeply staining 
nuclei and very little cytoplasm. No evidence of neoplastic 
epithelial or mixed tumor elements can be found. This tissue 
infiltrates between the tubules much as do the cells in leukemic 
infiltrations, and the capsule is also invaded. The diffuseness 
of this infiltrative growth and its equal involvement of both 
kidneys makes it resemble a leukemia or pseudoleukemia, but 
this diagnosis is untenable in view of the lack of involvement of 
other organs or lymph-nodes. The diagnosis of sarcoma is made 
largely by exclusion. 

' We wish to quote here a statement of the criteria used in our consideration 
of sarcoma throughout this work as expressed in our paper on Primary Spontane- 
ous Sarcoma in Mice (Jour. Cancer Res., 1917, ii, 1). "We recognize fully the 
difficulties that attend the differentiation of sarcoma, and for the purpose of 
this study have excluded every form of new growth concerning the nature of 
which there seemed any possible room for question. Therefore, we have not 
included numerous cases in which we think that the growths are probably 
sarcomatous, and many more in which we cannot be sure that the neoplasm 
is not sarcoma. On the other hand, the statistical value of our figures is 
lessened by the fact that we have imdoubtedly omitted some growths that are 
true sarcomas. Our figures represent minimal values only. From the standpoint 
of investigations in heredity, with which our work is particularly concerned, 
it is just as undesirable to call a sarcoma something else as to include a granuloma 
among the sarcomas, and hence the rigid classification adopted in this study of 
sarcomas is no more satisfactory for our heredity statistics than would be a lax 
classification that included some growths of doubtful nature. Therefore, in 
charting the heredity statistics it is necessary to recognize the absence of positive 
criteria for the differentiation of sarcoma, and to admit the borderline cases with 
a mark of interrogation to indicate this fallibility." 



TUMORS IN KIDNEY AND ADRENAL OP MICE 317 

7. Sarcoma of kidney. Old male mouse (24979), with the 
right kidney entirely replaced by a tumor, 16 x 12 x 10 mm., 
which infiltrated the adjacent tissues, including the ureter and 
the pelvis of the opposite kidney. The regional lymph-nodes 
were also invaded, and there was a mass in the root of the mesen- 
tery about 20 mm. in diameter, which seemed also to infiltrate 
the pancreas. 

Microscopically, the right kidney is foimd to be almost com- 
pletely replaced by a growth of large round cells, uniform in 
size, with very little cytoplasm, but with nuclei much larger 
than those of lymphocytes. There is no tendency to structural 
arrangement, the tumor cells infiltrating freely the tissues about 
the kidney, including the adjacent muscles and the pelvis of the 
opposite kidney, the ureter of which is surrounded by a mass of 
tumor. The regional lymph-nodes are replaced by tissue of the 
same character, but the mesenteric mass is completely necrotic, 
with few cells resembUng those of the tumor. By virtue of its 
highly infiltrative character and the large size of the cells, this 
seems to be an undoubted sarcoma. The other tissues showed 
no evidence of either leukemia or pseudoleukemia. 

8. Sarcoma of the kidney. Female mouse (26867) with much 
sarcosporidiosis, showed a marked enlargement of the left kidney, 
which was between two and three times the normal size, and 
infiltrated diffusely with a pink, fleshy tissue. Elsewhere in the 
body there were no important changes. Microscopically it is 
found that the enlarged kidney is infiltrated extensively, nearly 
all the renal elements being replaced by a tissue composed of 
round and elongated cells, considerably larger than lymphocytes 
and with more cytoplasm. They tend to be arranged in wide 
bands, growing out from the blood-vessels, but this arrangement 
is not constant or well developed. The neoplastic tissue grows 
out from the pelvis and tends to invade the adjacent tissues and 
the hilum of the opposite kidney. No similar tissue is to be 
found elsewhere in the mouse. The probable diagnosis is pri- 
mary sarcoma of the kidney. 

9. Sarcoma of the kidney. Female mouse (27148), with the 
right kidney much enlarged (18 x 15 x 32 mm.) and consisting of 



318 MAUD SLYE, H. F. HOLMES AND H. G. WELLS 

a firm tissue overlaid with softer portions. Microscopically it is 
found that the kidney is diffusely infiltrated with a growth con- 
sisting of polyhedral cells, smaller than epithelial cells usually 
are, but with a little more cytoplasm than lymphoid cells. These 
cells exhibit no structural organization and have replaced all but 
a few of the original renal elements. About one-third of the 
tissue is necrotic. A small amount of the neoplastic tissue 
infiltrates the hilum of the left kidney, which also shows amy- 
loidosis. No similar tissue is found elsewhere in the body, and 
there are no evidences of diffuse lymphoid hyperplasia or similar 
conditions. The lung contains a small benign adenoma. 

The tumor shows much variation in the size of the nuclei, 
hyperchromatism being common, and occasional very large cells 
are seen. Unfortunately there is too much post-mortem change 
for exact details to be seen. The general character of the growth 
and the absence of similar changes in other tissues make a diag- 
nosis of sarcoma seem inevitable. 

10. Bilateral sarcoma of kidneys, with metastasis in spleen. 
Female mouse (396) , with the uterus enormously distended with 
fluid because of vaginal obstruction by a seminal mass, and with 
great enlargement of both kidneys by a uniform white tissue. 
The right kidney measured 18 x 13 mm., the left 15 x 13 mm. 
On the anterior surface of the. left kidney is a 'depressed scar. 
In the spleen there is a tumor nodule, 4x5 mm. No other tumor 
growths were found. The adrenals were normal. 

Microscopically, both kidneys are found to be infiltrated dif- 
fusely, with replacement of 80 to 90 per cent of the renal tissue 
by a growth uniformly composed of small oval cells in solid 
masses, without any particular organization. No similar tissue 
can be found in any other organ except the spleen, which has a 
tumor nodule of the same structure as the renal tumor. The 
adrenals are in contact with, but not invaded by tumor. The 
tumor forms large masses in the hilum of one of the kidneys, or 
within the kidney itself. The left kidney is more involved than 
the right, and probably was the starting point of the sarcoma. 

11. Lymphosarcoma of kidney. Female mouse (13124), with 
two separate mammary carcinomas, and in the left kidney a 



TUMORS IN KIDNEY AND ADRENAL OF MICE 319 

nearly spherical mass about 14 mm. in diameter, which was 
partly necrotic. Microscopically this growth is composed of a 
dense mass of small round cells without visible cytoplasm, 
packed closely together without attempt at formation of any 
structure. It infiltrates the adjacent renal tissue freely, and there 
is a small amount of infiltration of the hilum of the opposite 
kidney. The structure is of distinctly neoplastic character, and 
there is no similar infiltration of other organs, or lymph-node 
enlargement, such as characterizes pseudoleukemia. The diag- 
nosis of sarcoma, in structure corresponding to lymphosarcoma, 
is the only one that can be made on the features presented. 

12. Lymphosarcoma of kidney. This seems to be similar to 
case 11, but unfortunately post-mortem decomposition has ad- 
vanced so far that accurate microscopic study is not possible. 
Such tissue as remains stainable resembles a lymphosarcoma. 
The mouse was a female (12533) with no other tumors, and the 
right kidney was replaced by a soft pink mass measuring 14 x 12 
X 12 mm. A small amount of the same tissue invaded the hilum 
of the left kidney. 

MESOTHELIOMA OF THE KIDNEY 

In our previous papers on tumors in the ovary (49) and testicle 
(50) of mice we have called attention to the not infrequent 
occurrence of tumors composed of polyhedral cells, presenting 
some characteristics resembling carcinoma and some resembling 
sarcoma, (figs. 4, 6, 7 and 8) and hence suitably designated meso- 
thelioma, in accordance with Adami. These tumors are charac- 
teristic of the urogenital anlage, and hence it is not surprising 
that growths of the same structure are found in the adrenal and 
kidney. We have observed the following cases of renal tumors 
that seem to belong to this group. 

13. This is a remarkable case in that a female mouse (21663), 
when but one month old exhibited two independent carcinomas of 
the mammary gland, and osteosarcomatous growths in the 
spinal column near the pelvis and in the left fifth rib. It lived 
but eighteen days more and at autopsy showed its left kidney 
also almost completely replaced by a tumor measuring 18 x 12 x 



320 MAUD SLYE, H. F. HOLMES AND H. G. WELLS 

10 mm. ; the right kidney, similarly involved, measured 10 x 6 x 6 
mm. (fig. 4). These renal growths are entirely different in 
structure from the other four tumors. They are alike and vary 
in appearance in different parts, some portions resemblii^ 



Fio. 4. Mesothelioma of Kid.vet 
This portion of the growth presents a sarcomatoua character; other portions 
exhibit more resemblaoce to epithelial growth. The tumor was found in a mouse 
but one month old, with four other tumors, which if not congenital must have 
developed very soon afUr birth. Mouse 21663. X 110. 

spindle-cell sarcoma while others are composed of lai^er, more 
polyhedral cells, arranged in a somewhat alveolar fashion, often 
separated by highly vascular septa. This structure corresponds 
to the type of growth often seen in tumors of the ovaries, testicles, 



TUMORS IN KIDNEY AND ADRENAL OF MICE 321 

and adrenals in mice, and agrees with the tumors called meso- 
thelioma by Adami and WooUey. Since, except for this case, 
there have been very few other cases of malignant tumors 
arising in mice less than six months of age, the occurrence in so 
young an animal of at least four independent primary growths 
representing three distinct types of malignant neoplasm, is a 
most remarable condition, without, as far as we know, a parallel 
in either mouse cancers or in those of any other animal. 

14. Male mouse (10011), with no other lesions of importance, 
had its right kidney largely replaced by a whitish tumor, forming 
a mass measuring 14 x 12 x 10 mm. Microscopically the tumor 
has largely replaced the kidney and has infiltrated the capsule in 
places. It is composed of small cells with a dark, round nucleus 
and a small amount of cytoplasm. These cells tend to form 
bands or pseudo-alveoli in some places, but for the most part the 
growth is composed of cells in a structureless mass. The left 
kidney was not involved by the tumor. 

15. This se^ms to be similar to case 14. Male mouse (9779) 
had an enlarged left kidney without other lesions of note. Micro- 
scopically the kidney is largely replaced by a mass composed of 
small cells slightly larger than lymphocytes and with more 
cytoplasm, showing no structural arrangement. The main neo- 
plastic mass lies at one side of the kidney, which it infiltrates 
slightly. The kidney itself shows some foci of round-cell in- 
filtration. Some of these areas sUghtly resemble the tumor, but 
probably are not a part of it. There are some small areas of 
calcification, some scars, and numerous hyaline casts. Post- 
mortem changes are too advanced for more accurate study. 
The opposite kidney contains no tumor, but there is the same 
amount of amyloid and calcification. 

SARCOMA OF RENAL PELVIS 

We have excluded numerous cases in. which a retroperitoneal 
tumor of sarcomatous character has invaded the kidney hilum, 
but there is one case in which both the gross and microscopical 
findings distinctly indicate that the tumor had its origin in the 
tissues of the pelvis itself. 



322 MAUD SLYE, H. F. HOLMES AND H. G. WELLS 

Female mouse (348) had a pale swelling extending downward 
and inward from the pelvis, composed of tissue of about the same 
consistency as the kidney and of uniform yellow color. On cross 
section the pelvis of the kidney was in the center of the mass 
formed by the tumor and the kidney, which measures 12 x 9 mm. 
The left kidney was of normal size with a pale area of swelling 
on the anterior surface. No other changes of importance were 
found except severe edema of the lungs. 

Microscopically the growth is found to lie symmetrically 
about the pelvis of the kidney and the upper end of the ureter, 
invading the wall of these structures extensively. From here it 
passes along the vessels deeply into the kidney, but does not 
extend far into the cortex. The kidney is about one half as large 
as the tumor, the tubules being much dilated and the glomeruli 
more or less hyalinized. The tumor also extends some distance 
along the renal capsule as a thin layer. As no growths are found 
elsewhere it is evident that this growth arose in the tissues about 
the hilum of the right kidney. There is only a small amount of 
invasion of the left kidney hilum. The growth is composed of 
large round cells with a delicate reticulum. There are several 
delicate vessels packed with lymphoid cells, but no other evi- 
dences of lymphatic origin in this tumor. The cells of the tumor 
are polymorphous with considerable cytoplasm, and deeply 
staining nuclei much larger than those of the ordinary lymph- 
cell, and usually larger than the nuclei of the renal epithelium. 
No mitoses are seen. No changes of importance are found in the 
other organs. 

This tumor is much less complex in structure than the cases of 
sarcoma of the renal pelvis in children described by de Vecchi 
and Salomon (51). 

ADRENAL TUMORS 

As mentioned aboye, no cases of adrenal tumors have hitherto 
been described in mice. We are able to report a few unquestion- 
able instances of such tumors. Despite the frequency of benign 
cortical adenomas of the adrenal in man, we have found but one 
such tumor in mice. It may be recalled that adrenal tumors 
have not been found frequently in other species of animals. 



TUMORS IN KIDNEY AND ADRENAL OF MICE 323 

Cortical adenoma of misplaced inter-renal adrenal rest (fig. 5). 
Between the kidneys of a female mouse (1921), with marked 
amyloidosis and chronic nephritis, was found a spherical mass 
about 5 mm. in diameter, attached neither to the kidneys nor to 
the intestines. Both adrenals were present at their normal site 



Fia. 5. Cohtical Adrenal Adenoma 
The tumor developed in an inter-renal rest of adrenal cortex tissue. A small 
zone of compressed adrenal tissue is seen at one side. Mouse 1921. X225. 

and were of normal structure and size, except for some peripheral 
round-cell infiltration. 

The tumor nodule is composed of a solid mass of cells resemb- 
hng those of the adrenal cortex except in lack of orderly arrange- 
ment, closely packed together, and flattening out a thin shell of 



324 MAUD SLYE, H. F. HOLMES AND H. G. WELLS 

adrenal cortex, evidently all that remains of an adrenal rest, since 
no medullary elements are to be found. In all respects this 
tumor corresponds to the simple adenoma of the adrenal cortex 
seen in man. 

MESOTHELIOMA OF THE ADRENAL 

This seems to be the commonest tumor of the adrenal, as also 
of the testicle, of mice, and it is quite impossible to distinguish 
on the basis of microscopic appearance between the mesothe- 
liomas arising in the different organs derived from the urogenital 
anlage. The cases in which the diagnosis seems certain are the 
following : 

1. Male mouse (10390) had in place of the right adrenal a 
spherical mass 5 mm. in diameter, which is completely encapsu- 
lated and does not involve the kidney. No other nodules or 
findings of importance elsewhere. Microscopically the nodule 
contains no remains of adrenal tissue, but consists of a solid timior 
made up of masses of large cells with considerable cytoplasm and 
large oval or spherical nuclei, arranged in atypical alveoli or 
broad bands with a very small amount of stroma containing 
thin-walled blood-vessels. In numerous places the capsule is 
infiltrated with tumor cells and there is some invasion of the 
areolar tissues about the adrenal, but no invasion of the adjacent 
kidney. This seems to be a typical mesothelioma of the adrenal 
in an early stage of malignancy. 

2. Mesothelioma of adrenal with peritonea^ metastasis. Fe- 
male mouse (12744), with abdomen greatly distended by a bloody 
exudate, presented at the site of the left adrenal a whitish mass 
about the size of the kidney. All through the abdominal cavity 
are masses of partially necrotic whitish tissue, especially attached 
to the liver and uterus, which do not seem to be involved by this 
growth. There are enlarged retroperitoneal lymph-nodes, but 
elsewhere no lymphatic involvement. There are no tumors in 
the lungs. 

Microscopically the tumor shows the usual features of the 
typical mesotheliomas, as described in the other cases, with slight 
tendency to alveolar arrangement. It does not seem to have 



TUMORS IN KIDNEY AND ADRENAL OF MICE 325 

infiltrated or produced metastases in the viscera. The right 
adrenal is surrounded, but not invaded, by the tumor and seems 
nonnal. The retroperitoneal nodes are replaced entirely by 
tumor tissue. 



Fia. 6. Mesothelioma of Adrenal 
This growth surrounded the lymph nodes and wob widely disseminated through 
the perirenal tissues. Mouse 7690. X 110. 

3. Bilateral malignant mesothelioma of adrenals (fig. 6). 
Female mouse (7699) with much subcutaneous edema, showed 
numerous enlai^ed subcutaneous lymph-nodes, some red and 
some pale, the largest measuring as much as 8 x 10 mm. There 
was a milky fluid in the peritoneal cavity, presumably because 
of pressure of enlarged lymph-nodes on the thoracic duct, for 



326 MAUD SLYE, H. F. HOLMES AND H. G. WELLS 

the retroperitoneal nodes were greatly enlarged, up to 6 x 12 mm. 
The mesenteric nodes were also enormously enlarged, one mass 
at the root of the mesentery measuring 40 x 20 x 20 mm. Both 
adrenals were greatly enlarged, each being about the size of the 
kidney, which was adherent to the adrenal but not enlarged or 
infiltrated by neoplastic tissue. The uterus seemed to be in- 
filtrated by tumor. The liver was not affected, and the lungs 
showed only a singte small nodule, although there was some tissue 
incfease in the upper mediastinum; both lungs showed a bloody 
edema, and there was a bilateral hydrothorax. 

Although the gross appearance suggested a general lympho- 
sarcomatosis or pseudoleukemia, yet the microscopic structure 
is of an entirely different character. Everywhere the neoplastic 
tissue presents the same appearance, being composed of a solid 
growth of cells with considerable cytoplasm, so that they look 
much like epithelium. The nuclei, which vary greatly in size, 
are much more solid than those of epithelial cells usually are. 
Mitotic figures are abundant. The invaded lymph-nodes are 
largely replaced by tumor cells, which are also found in the 
lymph- vessels of the lungs, but not in the liver. Both adrenals 
seem to be entirely replaced by the tumor, which lies upon the 
capsule of the adjacent kidney without any invasion of this 
organ. There are some areas of hemorrhage, but not much 
necrosis. The cells show no attempt at definite arrangement, 
but simply form a solid mass with numerous, poorly defined blood 
spaces. 

This tumor resembles in structure other growths found arising 
in the urogenital anlage, and in view of this and of the complete 
replacement of both adrenals it is most probable that it did arise 
in these tissues, although the extensive lymph-node involvement 
is unusual in adrenal tumors. The structure is not essentially 
dissimilar to that of the adrenal tumor 10390, except for the 
amount of extension. 

UROGENITAL MESOTHELIOMAS OF UNCERTAIN ORIGIN 

Because of the identity in appearance of mesotheliomas from 
all organs originating in the urogenital anlage, it is not always 



TUMORS IN KIDNEY AND ADRENAL OF MICE 327 

possible to decide the place of origin of some tumors which in- 
volve two or more of these organs, as shown by the following 
cases. 

12307. Mesothelioma of either adrenal or ovary. It is not 
possible to determine the origin of this tumor, which we have 
described in our paper on tumors of the ovary in the following 
words : 

The abdominal cavity shows several nodules whose exact origin is 
difficult to determine as the mate has partly devoured the body. The 
right ovary, is, however, easily distinguished. It measures 18 x 
12 X 12 mm. What seems to be the left ovary is 10 x 8 x 6 mm. 
There are 8 other similar nodules in the abdominal cavity, one being 
in the position of the left adrenal, measuring 10 x 8 x 8 mm. The 
other nodules are apparently in the mesentery. One lobe of the liver 
is converted into a tumor nodlile 14 x 10 x 18 mm., irregular and 
lumpy in outline, pink in color. 

The tumor shows everywhere the same structure, consisting of 
irregular alveoli composed of large cells with abundant cytoplasm with 
well defined borders and deeply staining nuclei. Mitotic figures are 
numerous. The character is that usual to mesothelial growths. The 
ovary cannot be positively identified, but one mass exhibited in the 
capsule structure suggests compressed ovarian tissue with degener- 
ated ova. In all respects this timior is identical with the malignant 
ovarian tumors just described. 

It seems probable that this tumor arose in the ovary which exhibited 
the largest growth, but it is not possible to exclude the adrenal as the 
primary site. 

The malignant tumors of the adrenal, ovary, and testicle com- 
monly exhibit identically the same histological picture as that 
seen in this case. 

We have also observed two other cases, described in the paper 
on ovarian tumors in which we cannot state whether the renal 
growth was primary or secondary. 

12876. The left kidney contained a mass of pink, fleshy 
tissue, 18 X 14 X 14 mm. The right kidney, which was slightly 
enlarged, contained no tumor. The right ovary consisted of a 
pinkish tissue resembling that in the kidney, and measured 



328 MAUD SLYE, H. F. HOLMES AND H. G. WELLS 

12 X 8 X 8 mm. In the mesentery was a similar, slightly paler 
mass, 16 X 8 X 8 mm. The retroperitoneal and subcutaneous 
nodes were not enlarged and no nodules were found in the limgs. 

Microscopically the tumor is alike in all three j>laces, consisting 
of a diffuse infiltrating growth of large roimd cells, which also 
invade the connective tissues about the kidney and ovary. It 
does not at all resemble the typical ovarian tumors, being ap- 
parently a round-cell sarcoma. We have no way of telling which 
of the three tumors was primary. The next case presents similar 
difficulties. 

26. This mouse had a tunwr mass about 8 x 10 mm. in the 
upper portion of the liver, with other smaller nodules near it. 
A similar small nodule was found in the right kidney. The right 
ovary was enlarged to two-thirds the size of a kidney, and was 
solid. Microscopically all these growths are composed of round 
cells, apparently a round-cell sarcoma. It is impossible to say 
which growth was primary. 

Among several tumors arising in the retroperitoneal tissues, 
mostly of sarcomatous type, are a few of a structure identical with 
the characteristic mesothelioma type of growth that arises in 
the urogenital anlage. In the two cases described below this 
character of growth was so marked that it seems probable that 
the tumors have arisen in some misplaced embryonic rest, since 
the organs of this series were distinctly not the starting point of 
the growths. Numerous tumors of this sort have been described 
in man. 

(22380) Malignant retroperitoneal mesothelioma with exten- 
sion through the body wall. Male mouse presenting externally 
a subcutaneous tiunor, involving the left hip and extending 
to the anus, the external measurements being 30 x 25 x 25 mm. 
When the body was opened it was found that this mass was an 
extension of a huge mass (40 x 20 x 18 mm.) which had sur- 
rounded the left kidney and pushed it to the ventral midline, 
invaded the spleen and pushed it mostly to the right of the mid- 
Une by formation of a mass that measured 20 x 15 x 13 mm., and 
extended downward through the pelvic cavity where it surrounded 
the rectum and seminal vesicles and broke out through the 



TUMORS IN KIDNEY AND ADRENAL OF MICE 329 

body wall. There was slight infiltration of the lower pole of the 
right kidney. The liver was slightly enlarged and mottled with 
areas of infiltration. There was some tumor in the lower part of 
the mediastinum, and the lungs show many small spots of neo- 
plastic tissue. There seemed to be no involvement of the 



Fig. 7. Secondary Mesothblioha in Liver 

The primary growth was either in the adrenal or in a retroperitoneal embryonic 

rest; it infiltrated the body wall and retroperitoneal tissues, and produced iuDU- 

merable metastases in the lungs and liver. This section shows the maaaes of tumor 

cells compressing and replacing the larger liver cells. Mouse 22380. XllO. 

lymph-nodes, either subcutaneous or abdominal. The testicles 
were not involved. 

Microscopically the tumor everywhere consists of masses of 
large cells with much deeply stainii^ cytoplasm, somewhat 
resembling liver cells, with a slight tendency to arrangement in 



330 MAUD SLYB, B. F. HOLMES AND H. G. WELLS 

cords or bands. Where the tumor invades the hver it can be 
seen that the cells are a little smaller and paler than the liver 
cells (fig. 7). 

The growth infiltrates the liver widely, invades the muscle 
of the body wall, invades the renal capsule but not the kidn^, 



FiQ. 8. Secondary Mesothelioma, op Lunos 
From the same case as figure 7. The section shows the extent of involvement 
of the lung. Mouse 223S0. X 60. 

infiltrates the abdominal sympathetic ganglia, surrounds densely 
the rectum and spreads into the seminal vesicle, invades the 
spleen, and in the lung appears as multiple timior cell emboli 
within the vessels and as large tumor nodules largely replacing 
portions of the lung (fig. 8). The left adrenal was found entirely 
embedded in but not infiltrated by the tumor. 



TUMORS IN KIDNEY AND ADHENAL OF MICE 331 

The microscopic appearances are identical with those usual 
in malignant adrenal tumors, but the adrenal was not involved; 
and in view of the retroperitoneal origin, the growth may be 
presumed to have arisen in an embryonal rest of the urogenital 
anlage. 

(9979) Malignant retroperitoneal mesothelioma infiltrating kid- 
neys. Small female mouse with extensive infiltration of the right 
thigh by a typical spindle-cell sarcoma, and with a mass above 
the left kidney, about as large as the kidney itself, which was 
infiltrated by the growth. The entire mass measured 20 x 14 x 10 
mm. The right kidney was not so much enlarged. The growth 
seemed to have arisen at the site of the left adrenal, and to have 
pushed the kidney forward. The liver was invaded by tumor 
and enlarged. No metastases were found elsewhere. Micro- 
scopically the tumor about the kidney bears no resemblance to 
the spindle-cell sarcoma of the thigh, being composed of a mass 
of polyhedral cells arranged in soUd masses, with a slight tendency 
to form bands or cords. The nuclei are not much larger than 
those of lymphocytes, but the cells have much more cytoplasm. 
The chief mass lies above and behind the left kidney, which shows 
considerable invasion through the capsule and about the blood 
vessels. There is a similar invasion of the right kidney but much 
less tumor about it. The right adrenal cannot be located; the 
left is free from tmnor although there is a necrotic area between 
it and the kidney. The liver shows a large nodule of the same 
sort of tumor. 

SECONDARY TUMORS OF THE KIDNEYS AND ADRENALS 

The mouse kidneys seem to be extremely insusceptible to 
metastatic invasion by tumors. In our entire series of primary 
tumors of mice, now in the neighborhood of 5000 cases, of 
which the predominating form is carcinoma of the mammary 
gland, we have never met with a metastatic growth from 
one of these tumors into the kidneys, with one possible 
exception, despite the frequency of pulmonary metastates, 
which often replace most of the lung. The only metastatic 
carcinomas of the kidney that we have seen have been found 



332 MAUD SLYE, H. F. HOLMES AND H. G. WELLS 

in four cases of primary carcinoma in the lung (3098, 11777, 
12373, 14242). A photograph of the first of these four cases 
appears in our paper on lung tumor (52). These carcinomas of 
the lung are especially likely to produce extensive metastases, 
and the suggestion of some authors that these pulmonary growths 
in mice should not be included among the true tumors is evidently 
based on ignorance of their character. The fact that they have 
furnished our only instances of metastatic carcinoma in the 
kidney is sufl&cient evidence of their true neoplastic character 
and the high malignancy of some types. 

Although we have seen many instances of widespread metas- 
tasis in sarcoma, the kidneys are almost immune from vascular 
sarcoma emboUsm. In our series of 87 cases of sarcoma in mice, 
although 23 showed metastasis but one hematogenous secondary 
nodule was found in the kidney, this coming from a mediastinal 
sarcoma (11791). Since then we have seen one other case of 
vascular metastasis of sarcoma into the kidney, from a sarcoma 
of the uterus (12058), with a large nodule of the same structure 
almost replacing the lower pole of the right kidney, and metas- 
tasis in the right ovary. Even the widespread growths of small 
round cells, which resemble the condition called lymphosarco- 
matosis in man, seem to aflfect the kidney but Uttle, for among a 
considerable number of such cases we have but one with a dis- 
tinct metastatic nodule in the kidney (7572), although several 
cases show extension from retroperitoneal metastases into the 
capsule, and then into the kidney. On the other hand, the 
kidneys exhibit extensive diffuse infiltration in leukemia and 
massive perivascular growths in pseudoleukemia in mice as in 
man. 

The malignant retroperitoneal growths, most of which resemble 
lymphosarcoma, and the malignant tumors derived from the 
adrenals, commonly invade the kidney by direct extension into 
the hilmn, often very extensively. The sarcomas primary in 
one kidney also tend to spread into the hilum of the opposite 
kidney. This ready invasion of the hilum of the kidney is a 
point of similarity of human and mouse neoplasms. 



TUMORS IN KIDNBT AND ADRENAL OF MICE 333 

No instance of tumor metastasis into the adrenal has ever 
been observed, except possibly in the few cases of widespread 
mesotheliomatous growths invading both the ovaries and adrenals, 
the origin of which is uncertain. 

SUMMARY 

In a series of 33,000 autopsies on mice of the Slye stock, dying 
natural deaths at all ages, but as far as possible living out their 
natural span of life, there have been observed the following cases 
of true primary neoplasm arising from renal or adrenal tissues: 
First, from the kidney, 16 tumors, classified as follows: 1 carci- 
noma, 3 adenomas, 1 hypernephroma, 7 sarcomas, 3 mesothe- 
liomas, and 1 sarcoma of the renal pelvis. Second, from the 
adrenal, 4 tumors, as follows: 1 cortical adenoma from a mis- 
placed inter-renal adrenal rest, 3 mesothelial tumors. Third, 
five cases of tumors of the mesothelial structure characteristic 
of urogenital anlage neoplasms, but of which the exact origin 
could not be determined because of their widespread growth 
at the time of death. As these 25 tumors occiured in 33,000 
mice presenting not far from 5000 other tumors, they are evi- 
dently uncommon tiunors of mice, at least in this particular 
stock. 

It will be noted that in this series there has been no instance 
of a mixed renal tumor of the Wilms tjrpe, which is so common a 
type of renal tmnor in man and apparently also in swine. Al- 
though inflammatory conditions are very prevalent in the kid- 
neys of mice, epithelial tumors are rare, and especially to be 
noted is the absence of even a single case of typical malignant 
hypernephroma, although one benign growth of this type was 
found. No epithelial tiunors of the renal pelvis were found, 
although there was one case of sarcoma that seemed to take its 
origin in the pelvis. 

Several instances of malignant retroperitoneal tmnors have 
been observed, mostly of sarcomatous structure, which usually 
invade the kidney. These have not been included in this series, 
except two cases in which the structure resembled that of the 

TBI JOUBHAL OW OAlfCBB XaSBABCH, TOL. TI, NO. 4 



334 MAUD SLYEy H. F. HOLMES AND H. O. WELLB 

mesotheliomas, suggesting that the tumor had its origin in mis- 
placed rests of the urogenital anlage. 

Secondary tmnors have never been found in the adrenals, 
and but rarely in the kidneys. Although this series includes at 
least 3000 cases of mammary carcinoma, often with widespread 
metastases in the lungs, we have never seen a secondary carcino- 
matous growth in the kidney. The only secondary carcinomas 
of the kidney as yet observed are fouT cases in which the 
primary carcinoma was in the lung, thus establishing the true 
neoplastic nature of these lung growths. In but two cases have 
metastatic sarcomas been seen in the kidney, if we exclude the 
numerous cases of invasion of the kidney by direct extension 
from para-renal growths. 

As to sex : In the entire group of renal and adrenal tumors, we 
have equal numbers in males and females, agreeing with the 
observation made on other timiors in mice that, in tumors not 
peculiar to the sex glands, there is usually little difference in the 
incidence in the two sexes. 

Differing from the tmnors previously studied, coincidence of 
other tmnors with the renal and adrenal tumors is imcommon. 
One mouse in this series had a spindle-cell sarcoma of the thigh. 
One mouse had a small, benign papillary adenoma of the lung. 
Only two mice had a mammary carcinoma, and one of these 
(21663) was a remarkable case, for this animal, when but one 
month old, was found with two independent mammary carcino- 
mas, and with osteosarcomas in the spinal column and in a rib. 
It lived eighteen days more, and at autopsy there was also found 
a mesotheliomatous type of growth involving both kidneys. 
Except for this imique case there have been practically no in- 
stances of malignant timiors in mice less than four months of 
age, and few imder six months. Most of the renal sarcomas 
occurred between the ages of seven months and one year, which is 
somewhat earlier than the usual time of appearance of epithelial 
growths; this, of course, corresponds to experience with human 
neoplasms. 

The epithelial renal and adrenal tumors furnished no illustra- 
tion of metastasis, but in three cases of sarcomatous or meso- 



TUMORS IK KIDNEY AND ADRENAL OF MICE 335 

theliomatous growths there was noted involvement of the ad- 
jacent lymph-nodes; in two there were pulmonary, in two hepatic, 
and in one splenic metastasis, and in one case there were numerous 
peritoneal growths. The mesothelial type of growths produced 
the most extensive metastasis and the most widespread infiltra- 
tion of the body wall. 

A review of the literature on renal tumors throughout the 
animal kingdom, which constitutes the introductory part of this 
article, discloses but six other cases of renal tumors in mice, 
all epithelial, and no other adrenal tumors. 

REFERENCES 

(1) Ttzzer: Jour. Med. Res., 1907, xvii, 166; 1909, xxl, 479. 

(2) Haaland: Fourth Sci. Report, Imperial Cancer Research Fund, London, 

1911, p. 1. 

(3) Murray: Third Sci. Report, Imperial Cancer Research Fund, London, 

1906,73. 

(4) Stumpf: Zieglers Beitrftge, 1910, xlvii, 571. 

(5) MgCot: Jour. Med. Res., 1909, xxi, 285. 

(6) WooLLET AND Whbrrt: Jout. Med. Res., 1911, n. s. xx, 205. 

(7) BxTLLocx AND Rohdbnburq: Jour. Cancer Res., 1917, ii, 39. 

(8) Bbatti : La Semana Medica, Buenos Aires, 1917, xxiv, 643. 

(9) Loewenstein: Beitr. s. klin. Chir., 1911, Ixxvi, 750. 

(10) MgCot: Jour. Infect. Dis., 1914, xiv, 53. 

(11) Fox: Jour. Path, and Bacteriol., 1912, xvii, 217. 

(12) CxTRTis: Jour. Agric. Res., 1915, v, 397. 

(13) Wernicke: Ztschr. f. Krebsforsch., 1911, x, 168. 

(14) BOrger: Ztschr. f. Krebsforsch., 1914, xiv, 526. 

(15) JoBST AND Ernebti: Ztschr. f. Krebsforsch., 1915, xv, 1. 

(16) White: Zool. Soc., Philadelphia, 46th Report, 1918, p. 28. 

(17) Fox: 48th Report, 1920; 49th Report, 1921. 

(18) Fox: 43rd Report, 1915; Jour. Path, and Bacteriol., 1912, xvii, 227. 

(19) Fox: Zool. Soc, Philadelphia, 42nd Report, 1914. 

(20) Fox: Jour. Path, and Bacteriol., 1912, xvii, 227. 

(21) Fox: Zool. Soc., Philadelphia, 49th Report, 1921. 

(22) Seliqman: Proc. London Zool. Soc, 1907, p. 173. 

(23) Baird: Jour. Cancer Res., 1917, ii, 103 

(24) BoRELL AND Masson: Bull, assoc. frauQaise pour T^tude du cancer, Paris, 

1911, iv, 172. 

(25) Scott: Jour. Cancer Res., 1917, ii, 367. 

(26) Sticker: Arch. f. klin. Chir., 1902, Ixv, 616. 

(27) McFadtean: Jour. Comp. Path, and Therap., 1899, xii, 139. 

(28) Kimura: Gann, 1917, xi, 38. 

(29) Trotter: Jour. Compt. Path, and Therap., 1911, xxiv, 1. 



336 MATJD SLYE, H. F. HOLMES AND H. G. WELLS 

(30) Stbinke: Frankfurter Ztschr. f. Path., 1910, v, 167. 

(31) RouBST AND Wolf: Ann. de m^., 1921, viii, 464. 

(32) Cadiot: Bull, de TAcad. de m6d., Paris, 1921, Ixxxv, 396. 

(33) RoFFo: Rev. d. Inst. BacterioL, Buenos Aires, 1918, i, 333. 

(34) Murray: Third Sci. Report, Imperial Cancer Research Fund, 1906, p. 41. 

(35) Hamburger: Virchows Arch. f. path. Anat., 1889, cxvii, 422. 

(36) White: Zool. Soc, Philadelphia, 46th Report, 1918. 

(37) Bland-Sutton: Jour. Anat. and Physiol., 1885, xix, 464. 

(38) Wells: Trs. Chicago Path. Soc, 1908, vii, 140. 

(39) Pick: Berl. Med. Gtosellsch., 1905, ii, 406. 

(40) Williams: Natural History of Cancer, p. 11. 

(41) Teutschlaender: Ztschr. f. Krebsforsch., 1920, xrii, 322. 

(42) Kimura: Gann, 1920, xiv, 10. 

(43) Steinkb: Frankfurter Ztschr. f. Path., 1910, ▼, 167. 

(44) Smallwood: Anat. Anz., 1905, xxvi, 652. 

(45) Fox: Jour. Path, and BacterioL, 1912, xvii, 217. 

(46) Fox: Zool. Soc, Philadelphia, 49th Report, 1921. 

(47) Slte, Holmes, and Wells: Jour. Cancer Res., 1920, ▼, 205. 

(48) Ewino: Neoplastic Diseases, 2d ed., Philadelphia 1922, p. 739. 

(49) Stlb, Holmes, and Wells: Jour. Cancer Res., 1920, v, 205. 

(50) Slte, Holmes, and Wells: Jour. Cancer Res., 1919, iv, 207. 

(51) DE Vegchi and Salmon: Quoted by Ewino: Neoplastic Diseases, 2d ed., 

Philadelphia, 1922, p. 767. 

(52) Slte, Holmes, and Wells: Jour. Med. Res., 1914, xxx, 417. 



A CRITICAL INVESTIGATION OF THE FREUND- 

KAMINER REACTION 

LOUIS HERLY 
From Columbia Universityt InstUtUe of Cancer Research, F. C. Wood, Director 

Received for publication May 24, 1922 
BIBLIOGRAPHY 

Human tumors 

• 

In 1910, E. Freund and G. Kaminer (1) described certain 
phenomena occurring when tumor cells are mixed with the 
blood serum from non-cancerous individuals, on the one hand, 
or blood serum from cancer patients on the other hand. This 
relationship became known as the Freund-Kaminer reaction. 
These investigators used tumors obtained at autopsy, because 
they found it harder to make cell emulsions with fresh material. 
The healthy portions of the tumor were minced, put into a cloth 
and pressed out by hand in water containing 0.6 per cent NaCl 
plus 1 per cent NaFl; the connective tissue and blood-vessels 
remained behind and the cells escaped. The resulting emulsion 
was centrifuged and the cells were then washed two or three 
times with 0.6 per cent saUne solution. The washed cells were 
finally suspended in an equal amount of 0.6 per cent saline solu- 
tion, and enough NaFl was added to make the solution 1 per 
cent NaFl; this proportion of NaFl will preserve the cells for 
three weeks. In many cases the cells agglutinated during 
washing, so that one washing had to suffice. Organ emulsions 
were prepared in the same way, and freed of blood by washing 
once with distilled water. 

The serum was obtained sometimes fresh, sometimes post- 
mortem; sometimes, in later experiments, one-tenth of its vol- 
ume of 0.5 per cent NaFl solution was added as a preservative. 

337 



338 LOUIS HERLT 

To perfonn the test, ten drops of serum and one drop of cell 
emulsion were mixed, and a drop of 0.5 per cent NaFl solution 
added ; the whole was then well mixed, and the cells in one drop 
of this emulsion were counted in a Thoma-Zeiss counting chamber. 
The mixture was then incubated in a test tube with a rubber top 
at from 37°C. to 40°C., and a second count made. When there 
were about 20 cells to a large square, their enumeration was 
easy; more than this were difficult to count, and a mixture 
with 40 cells to the square, which gave a negative result, gave a 
positive result with 20 cells to the square. 

Freund and Kaminer f oimd that the serum from noncancerous 
subjects destroys most of the cancer cells. The experiment was 
controlled by incubating an emulsion of cancer cells in a corre- 
sponding amount of 0.6 per cent saline solution containing 0.3 per 
cent NaFl; no diminution in the number of cells was found. 
Non-cancerous serum did not destroy normal cells, whether 
these were from normal persons or from those having cancer. 
Heating the serum to 55°C. abolished its power to destroy cancer 
cells. The authors concluded that the serum of non-cancerous 
individuals must contain a coctolabile substance with carcinoly- 
tic powers. 

In other emulsions similarly prepared, except that the serum 
came from cancer patients, there was no diminution in the cell 
content. All tests were made with the cancer cells of one 
patient and sera of others. 

Freund and Kaminer concluded with the question: ''Is the 
destruction of carcinoma cells by normal serum due to some 
property in the serum, or to some special fragility of the cancer 
cell itself?" and they argued that the resistance of carcinoma 
cells to carcinoma serum points to some lytic property in normal 
serum rather than to a vulnerability of the carcinoma cell. 

Freund and Kaminer's paper aroused great interest, because 
of the diagnostic possibilities of their reaction, and the experi- 
ment was repeated by many other investigators. Thus P. v. 
Monakow (2) f oimd lysis absent in 86 per cent of cases of car- 
cinoma. A cell destruction up to 10 per cent he considered 
negative; between 10 per cent to 25 per cent doubtful; and above 



INVESTIGATION OF FRET7ND-KAMINER REACTION 339 

25 per cent positive. He observed that besides cancer cells the 
emulsions contain connective tissue cells, which are not de- 
stroyed. In his hands only three-fourths of normal sera caused 
lysis of cancer cells; one quarter of the normal sera tested acted 
like cancer sera and one-fifth of all cancer sera destroyed cancer 
cells like normal sera. He, as well as Freund and Kraus, f oimd 
that the reaction often failed in sarcoma. 

Stammler (3) concluded that the Freund-Kaminer test showed 
80 per cent positive reactions. 

Ranzi (4), in 1911, in a review of diagnostic reactions for malig- 
nant tumors, concluded that the serum of cancer patients con- 
tains a substance protective for cancer cells. 

Kraus and v. Graff (5) attacked the problem from a different 
angle. They knew from the work of Salomon and Saxl that 
both in pregnancy and in carcinoma the excretion of oxjrprotein 
is increased, and that the polypeptids are increased in carcinoma 
and in the later months of pr^nancy. On the assumption of 
some analogy between the two conditions, they tested the serum 
of pregnant women and serum from the umbilical cord, and 
found that cancer serum as well as that from the umbilical cord 
failed to dissolve cancer cells. The serum of pregnancy caused 
'lysis of cancer cells only inconstantly, while serum drawn after 
confinement had no Ijrtic action. They concluded that the 
placenta produces alterations in the serum causing a loss of 
carcinolytic qualities or the formation of a protective substance. 
Rabbit and guinea pig serum acts as normal human serum, and 
rat, goat, and sheep serum act as carcinoma serum, the former 
being lytic and the latter not. 

Kraus, v. Graff, and Ranzi (6) also tested the reaction, and 
foimd that 71.4 per cent of tumor patients examined gave a 
positive reaction (lysis below 25 per cent) and 3.5 per cent a 
partial lysis. Of patients with other diseases and with benign 
tumors, 61.2 per cent gave a negative reaction, 15.3 per cent a 
positive reaction, and 23 per cent a partial lysis. Seven cases 
that had been operated on and were clinically free from re- 
currence gave a negative reaction (lysis over 50 per cent) ; hence 
they concluded that the Freund-Kaminer reaction is not based 



340 LOUIS HBRLY 

on an acquired tumor disposition but is due to changes in the 
blood serum depending upon metabolic changes in the tumor. 

Kraus and Ischiwara (7) tested embryonal cells against cancer 
cells, and found that human embryonal cells show even greater 
lysis with normal serum than do cancer cells. Furthermore, 
embryonal cells were cytolyzed by a cancer serum which had no 
action on cancer cells. There was no lysis with fetal blood serum ; 
but maternal (retroplacental) serum was Ijrtic. 

Rosenthal (8) continued along these general lines, using a 
25 per cent emulsion of cells suspended in physiological saline 
solution as a control. He beUeved that only a lysis exceeding 
25 per cent can be ascribed to the action of serum. His results 
were as follows: Fetal cells plus pregnancy serum gave a lysis 
beyond 25 per cent. Fetal cells plus fetal serum showed a 
lysis beyond 25 per cent. Fetal cells plus lunbilical cord serum 
gave no cjrtolysis. 

Arzt and Kerl (9) also have attempted to determine the 
diagnostic value of the Freimd-Kaminer reaction. They found 
in the group of tumors examined 83 per cent correct diagnoses 
and 17 per cent wrong. These diagnoses were verified by 
biopsy, by operation, or by autopsy. In healthy persons and in 
those sufifering from non-tumorous diseases the reaction gave * 
87.5 per cent correct and 12.5 per cent incorrect diagnoses. 
Freimd and Kaminer, when consulted about these figures, 
suggested that the wrong diagnoses might be due to such errors 
in technique as: (a) obtaining blood during the height of diges- 
tion; (&) admixture of red blood cells with the serum; (c) the use 
of insufficiently fresh serum; (d) the use of cells that had lost 
their sensitiveness. 

Kraus, Ischiwara, and Wintemitz (10) reached the decision 
that human serum can destroy cancer cells but not normal cells, 
and that embryonal cells behave as cancer cells in relation to 
adult and umbilical cord serum. 

This work attracted the attention of Coca (11) who carried out 
more than 150 experiments, some of them in the laboratory 
of Freund in Vienna. All his experiments, however, even 
those performed in Freimd's laboratory, resulted negatively, 



INVESTIGATION OF FREUND-KAMINER REACTION 341 

no cytolysis being observed in any instance with fresh noimal 
human, dog, or horse serum. He believed, therefore, that the 
cytolytic action of normal sera described by Freund and Kaminer 
depended upon some factor as yet uncontrollable, and that it 
cannot be made the basis of a differential test for malignant 
tumors. 

Freund and Kaminer have recently published (12) further 
experiments, to show that normal serum and tissues contain 
an organic fatty acid able to destroy cancer cells; this they call 
"normal acid." Cancer and the serum of cancer patients con- 
tain an acid that protects the cancer cell by destroying the 
normal acid. Pending an identification of the ''normal acid," 
an investigation was undertaken of the saturated dibasic fatty 
acids, the series to which they beUeve ''normal acid" belongs, 
in order to see which ones are able to destroy cancer cells. 

The most recent publication is that of Frankenthal (13), who 
f oimd lysis of cancer cells treated with normal serum to beyond 
50 per cent; but no lysis of cancer cells treated with the serum of 
fourteen patients having non-cancerous diseases; 25 sera of 
cancer patients gave 7 (28 per cent) negative results. Her con- 
clusions are: (a) normal serum destroys cancer cells, but also 
normal liver cells in lesser degree; (b) cancer serum acts differ- 
ently from normal serum, in that it protects the cells from de- 
struction up to a certain degree; (c) the Freund-Kaminer reaction 
is neither constant nor specific and, therefore, of no diagnostic 
value for the present. 

Animal tumors 

Freund and Kaminer used tumors (as stated above) that had 
been in the mortuary or the laboratory for indefinite periods of 
time, and it is open to speculation what proportion of the cancer 
cells so used were alive. They also state that the serum used 
in their experiments was sometimes obtained fresh and some- 
times postmortem, and one may justly ask whether this serum 
might not have been more or less decomposed. Hirschfeld and 
Ischiwara, working independently of each other, apparently saw 
this objection, for they used freshly drawn animal serum. 



342 LOUIS HERLY 

Animal tumors also have been employed, their great advantage 
being, of course that the viability of their cells can be tested by 
inoculation. Thus Hirschfeld (14) made a very fine emulsion 
of tumor cells, exposed it to the serum of non-tumor bearing 
animals or to that of rats with sarcoma in the proportion of 
1 to 6, incubated it for three hours, shaking the mixture every 
twenty minutes; and then diluted it with equal parts of 0.6 per 
cent saline. One cubic centimeter of this suspension was 
inoculated into the right inguinal region of normal animals, and 
a series of control animals was inoculated with tumor emulsion 
in physiological saline solution. The tumors in the animals 
inoculated with timior cells treated with normal serum were 
fewer and smaller than in those inoculated with tmnor cells 
treated with tumor serum. He concluded that this experiment 
proves again that normal serum has a damaging action on the 
vitality of tumor cells, while the serum of tumor-bearing animals 
lacks this quality. 

Finally, Ischiwara (15), working with rat sarcoma demon- 
strated that the cell reaction appears relatively late (thirty days) 
in the serum of rats, and has some relation to the size of the 
tumor. 

The reaction has been tested from the refractometric side by 
Koritschoner and Morgenstem (16), and by Koritschoner 
alone (17). 

EXPERIMENTAL 

In retesting Freimd and Kaminer's experimental conclusions 
the present writer has also retested Kraus and v. Graff's observa- 
tion (5) that the serum obtained from rabbits and guinea-pigs 
acts on cancer cells as does normal serum; i.e., causes lysis, 
whereas that from rats, goats, and sheep acts like carcinoma or 
placenta serum, i.e., has no appreciable Ijrtic effect. If this 
were proved to be correct, then the conclusions drawn from all 
the experiments in which normal rat serum was used (Hirschfeld 
and Ischiwara) would obviously have to be considered erroneous. 

In a series of experiments in which blood was drawn from the 
heart of rats and guinea-pigs, and from an ear vein of rabbits, 
this observation was tested in the following manner. 



INVESTIGATION OF FREUND-KAMINER REACTION 343 

The blood, drawn under aseptic precautions, was centrifuged 
and the resulting serum was pipetted off, care being taken not to 
allow any admixture of red blood cells, since Freund, in a dis- 
cussion of reported failures had attributed them to faulty tech- 
nique. As another suggested error was the use of serum not 
sufficiently fresh, all sera obtained were used immediately. 
Freund also attributed failure to the use of blood obtained at 
the height of digestion, and to avoid this the experiments now 
being described were made with animals which had not been 
recently fed. 

The tmnor used in making the suspension was the Flexner- 
Jobling rat carcinoma, since all other experiments along these 
lines had been carried out with rat sarcoma in spite of v. Mon- 
akow's observation, corroborated by Freund and Kaminer, 
that the reaction often fails with sarcoma. The tumors selected 
were from thirty to seventy days old, it being generally believed 
that transplantabiUty and growth capacity increase, in a general 
way, with the age of the tumoi*. In no case was the skin of the 
animal attached to the tumor, nor were any of the growths 
ulcerated. Furthermore if an animal appeared sick its tumor 
was not used. 

The tumors were extirpated under aseptic precautions and the 
connective tissue capsule was carefully removed, for previous 
observers had stated that the admixture of connective tissue 
vitiated the test. Thus the only connective tissue remaining 
was the small amount comprising the stroma. 

Only the healthy margin of each tmnor was used and micro- 
scopic sections were made for purposes of checking up each 
experiment. 

A fine emulsion of the tximor was then made by thoroughly 
mincing it with scissors, in order that the serum might come into 
contact with all parts of the growth. Various machines devised 
for emulsifying tumors were at first used but it was found that the 
resultant particles were too coarse and that a much finer emulsion 
could be obtained by using very sharp, curved manicure scissors. 
All manipulations, of course, were conducted under strictly 
aseptic precautions. 



344 LOUIS HEBLT 

This fine tumor emulsion was next added to the serum in the 
proportion of 1 to 6, and the suspension was placed in test tubes 
and incubated at 37°C., for three hours. Some fresh tmnor 
emulsion was immediately transplanted, without exposure to 
senmi, into a series of healthy laboratory rats known to be 
susceptible to the Mexner-Jobling carcinoma, in order to prove 
the viability of the tiunor employed. The amoimt transplanted 
was 0.003 gram fresh tumor emulsion, which was deposited in the 
right inguinal region with the usual inoculating needle. The 
tubes placed in the incubator were each supplied with a sterile 
glass rod and the emulsion was vigorously stured every half 
hour in order to bring all particles of the tumor emulsion thor- 
oughly into contact with the serum and so allow the serum to 
exert its influence on the largest possible number of tumor cells. 
At the end of the three-hour period all supernatant serum was 
poured ofif and the fine particles transplanted into a series of rats. 

The protocols of a few experiments selected at random from a 
considerable number follow: 

Comparison of the number and size of the tumors in figure 1 
shows no distinct difference between the control tumors and 
those exposed to rabbit serum. In the serum series there is 
foimd one receding tumor; but recession of the Flexner-Jobling 
carcinoma is a common event, and does not vitiate the results 
of this experiment. One inoculation was negative, but only 
about 80 per cent of inoculated animals, under routine conditions, 
grow this tumor. One animal died before the first charting. 
On the whole, one may say that the animals in both series had 
tumors of from 10 mm. to 40 mm. in diameter and that the takes 
are entirely comparable in both series. 

Reference to figure 2 shows that in both series after forty-five 
days there are tumors from 10 mm. to 40 mm. in diameter, the 
nimiber of growing timiors being practically 100 per cent, with 
the exception of one receding tumor in the serum series. 

It is evident from figure 3 that the tmnors were more numerous 
as well as larger in the serum series than in the controls. 

In the next group of experiments the effect of incubation in 
guinea-pig serum was studied. 



INVESTIGATION OF FBEUND-KAMINBB REACTION 345 

As may be seen from figure 4, 45 days following the inoculation 
there were 100 per cent takes in both series, and there is no 
obvious difference between the results in the two groups. 

miM PUIN TUMM MUUION 



« 17 i4 M M 4S 
• • • • t 


« 17 M M M 41 
«• « • • f • 

« • • • • • 


•»•§•# 


It ■ • • • # # 


t. . . « • • 




• • t •• 




• • • • • 




f . - - • • • 





10 . • t • • - 

tl - • - - - - 



Fio. 1. Experiment -tt-t" 

89 A 

An emulsion of the Flexner-Jobling rat carcinoma, diluted with five parts of 

rabbit serum, was incubated for three hours and then transplanted into a series 

of 12 rats (nos. 1 to 12). Fresh tumor emulsion was inoculated into a series of 

6 rats (nos. 13 to 17) to act as a control. 

Reference to figure 5 will prove that there are 80 per cent takes 
in the serum group, whereas in the fresh tumor, or control series, 
there are only 40 per cent takes, and the tumors are actually 
smaller. Clearly the guinea-pig serum exerted no deleterious 
effect on the tumor cells. 



346 LOUIS H£BLY 

It can be seen from figure 6 that forty-two days after the 
inoculation there were 100 per cent takes in both series and 
there is no obvious difference that can be observed in comparing 
the results in the two series. 

Rabbit and guinea-pig serum may have a different action on 
hmnan cells, but for the purpose of the present investigation the 
road is cleared by the proof that the serum from neither of these 
species has a deleterious effect on the tumor cells employed. 

ftAWIT SEnm MJUN TUMM EMIlSmi 

DiVf 
I0f7l43fdt4f IOf7l4Mdt4S 



• — ' 






10 CM 

Fig. 2. Experiment 



FRC 



89 E 

Five rats (nos. 1 to 5) were inoculated with an emulsion of the Flexner-Jobling 
rat carcinoma incubated in rabbit serum in the proportion of 1 to 6, and 5 rats 
(nos. 6 to 10) with fresh tumor emulsion. 

The results of these experiments do not agree with Kraus and 
V. Graff's observation that the serum obtained from rabbits and 
guinea-pigs exerts a damaging action on cancer cells; for, in 
some instances (figs. 3 and 5) the cancer cells subjected to the 
action of serum from these animals grew even better than did 
cells from the same tumor which had not been subjected to the 
action of the serum. We do not attempt to explain the reason 
why previous observers have obtained such different results. 

The objections to Freund and Kaminer's method have been 
already discussed. To obviate these, a series of experiments was 



INVESTIGATION OF FREUND-KAMINSR REACTION 



347 



carried out on animal tumors because here both cells and serum 
can be obtained perfectly fresh, and the viability of the tumor 
cell can be tested by inoculation. Cells from rapidly growing 
rat carcinomata (Flexner-Jobling) were suspended in the serum 



10 17 



1 • 



iABBir SEtUM 
24 31 38 4f tt ft 

• ••ft 





• • 



- — • 



• • • • • 



• • # 






fUm TVNN IMHUIM 

17 t4 31 3t 4t I 
• • • f • 




> • 



• • •• 



•• •• 



11 ^ 



— - -•» 



10 CM 



FlO. 3. EXPEBDODNT 



FRC 

90K 



A series of rats (nos. 1 to 7) was inoculated with an emulsion of the same tumor 
incubated in rabbit serum in the proportion of 1 to 6, and 6 rats (nos. 8 to 12) 
with fresh tumor emulsion to serve as a control. 



348 



LOUIS HERLT 



of normal rats, or of rats bearing this tumor, without the inter- 
vention of any extraneous substance such as NaCl or Nafl. 
It was realized, of course, that only the first half of the experi- 
ment is really valid, since the serum of an animal with a trans- 



ff 24 3f M 4i 
ft • • 



17 tf 



M 4S 





If 




# # 

* # 



• •• 



- - • • t 
• • • •• 



• • 



• • 



Fio. 4. ExPBsncsNT 



FRC 

89 A 



A series of 10 rats (nos. 1 to 10) was inoculated with an emulsion of the Flezner- 
Jobling rat carcinoma incubated in guinea-pig serum in the proportion of 1 to 6, 
and a series of 5 rats (nos. 11 to 15} with fresh tumor emulsion for controls. 

planted tumor may be entirely different from that of one with a 
spontaneous neoplasm. But the extreme rarity of carcinoma 
in the rat, and the impossibility of using a sarcoma left no other 
course open. Mouse carcinoma was not used because of the 
scanty amount of serum obtainable from so small an animal. 



mVEBTIOATION OF FBBUND-KAMINER SBACTION 



349 



8 
8 
8 
8 
S 

a 

§. 




I 

9 



I 

O 



I 
I 

I 






s 



« o 
.S "S 

^ 3 

.S S 

s a 
.a 8 



9 to 

IS 

rfj -^ 
^ OB 

.2 5 

3 S 



o 

»4 



§ 



8 
8 







a 



M 

o 
Em 



sg 

•^ o 



.a 



S 



5"S 

§•2 

" 8 
S o, 

S S 

'c a 

»4 






350 



LOUIS HERLY 



Nevertheless, it can be shown whether or not normal serum 
damages the tumor cell; if it does not, the distinction between 
the serum of a normal animal and of one with a spontaneous 
tumor vanishes. 



DAYS 
7 14 



I 



— 4 



21 28 3S 42 

• ••• 



7 M 21 



fl 






• • • 

• • •• 

• • • f 
« • t # 

• • • t 

• # • • 



11 

14 



17 



! 



It It 




♦ ? 




t« 




• • 




: 1 




♦ 1 




♦ f 


• 


• » 



mr 



Iff 



Fig. 6 Expebucent 



FRC 
92 B 



A series of rats (nos. 1 to 10) was inoculated with an emulsion of the Flex- 
ner-Jobling rat carcinoma incubated in guinea-pig serum in the proportion of 
1, to 6, and a series of 10 rats (nos. 11 to 20) with fresh tumor emulsion for 
controb. 

The experiments were carried out exactly as were those with 
rabbit and guinea-pig serum. The blood was obtained from the 
hearts of rats bearing Flexner-Jobling tumors from four to eight 
weeks old. All animals had large single or multiple tumors, 
ranging from 20 mm. to 35 mm. in diameter, none of which were 



INVSSTIQATION OF FBETJND-KAMTNER BBACTION 351 

fibKed to the skin or ulcerated. Nor did any of these animals 
manifest any evidence of iUness. 

While the other manipulations necessary to the experiment 
were conducted, the serum was placed in the refrigerator at 
8®C. Similarly a nimiber of rats without tmnors were bled 
and the serum also placed in the refrigerator. 

A healthy tumor was chosen, its capsule carefully stripped 
off, any necrotic material thoroughly scraped out, and the outer 
layer of the remarning healthy tumor tissue finely emulsified. 
All tiunors used in these experiments were from thirty to ninety 
days old and microscopic sections of each tumor used were made 
and filed away for purposes of identification. 

The sera were now taken from the refrigerator and to each was 
added this finely minced tumor in the proportion of 1 part of 
tumor to 5 parts of serum. A sterile glass stirring rod was added 
to each tube, and the tubes were incubated at 40^C. for three 
hours. Every half hoiu* each tube was stirred with its glass 
rod in order to bring all tumor cells repeatedly into contact with 
the serum. 

At the end of the incubation period all the serum was poured 
off and the fine tumor particles were transplanted into a series 
of rats in the manner previously described. A similar dose 
(0.003 gram) of fresh tumor emulsion was transplanted in the 
same way and at the same site without its having been exposed 
to any serum. 

The protocols of a few typical experiments follow: 

Figure 7 shows that the tumor cells incubated in normal serum 
grew better and more imiformly than did those incubated in 
tumor serum, and even better than the imtreated tumor emulsion 
which had not been in contact with any serum whatsoever. 

Figure 8 shows again that there was better growth in the 
series of animals inoculated with cancer cells incubated in normal 
serum, than in the tumor serum series. The results in the im- 
treated tumor emulsion series were similar to those obtained in 
the normal serum series. 

After thirty-eight days, as figure 9 shows, the results in all 
three series were practically alike, the tumors varying in diameter 
between 7 mm. and 20 mm. and growth being about the same. 



n . 



tt , 









f 



t 

t 

« . ; t t + 



I 



17 14 



• • + 



n t 



M t 



21 « 



I 






f 



— • 

23 I t 

24 -. . 



Ml 
tf 



HIHimMIMUiM 



n 17 M M » 

- • , , • 



17 - _ 






FiQ. 7 Experiment 



FRC 
83 B 



An emulsion diluted in the proportion of 1 to 6 with tumor serum, and in the 
same proportion with normal serum, was incubated and then transplanted into 
two series of animals, twelve each in number. Into a series of five animals a fresh 
undiluted tumor emulsion was transplanted. 

352 



■^-^i^\ 



INVESTIGATION OF FREUND-KAMINER REACTION 



353 



I 



I 



17 M « 31 


M IT M 3) 91 

.'. . lit 


• t •% 


u > * % %P 


• < « t 


» • » • « « 


. ♦ f < 


«*•§•• 


• • < ft 


17 • « • • • 


• • « • 


It ! ff f f # 


- • • • 


It • t t ! t 



II . 

+ 



- + 



u . 



PUINTUNN 

•V 17 

M « I 



17 



M M » 

» t 

* » * 



• « • + 



Fig. 8 Expxrimxnt 



FRC 

84N 



An emulsion diluted in proportion of 1 to 6 respectiyely with tumor serum and 
normal serum was incubated for three hours and then inoculated into two series 
of twelve animals each and the fresh tumor emulsion into a series of five animals. 



354 



LOXnS HBBLY 



TVMOII SEIHM 

DAYS 
10 17 24 31 3t 



NOIMAL SCIUM 



1 



t 

11 ^ 

12 — 









10 17 U 31 
13 - • < ^ 

M ■ • • • 



17 
It 
13 



11 - - _ 



• 



'It 
fit 



KAIN TUNOI CMUUION 
10 17 24 31 30 



30 - 



• 



t • 



27 • • 

20 - - 

29 



• 



— — — • 



10 CM 

Fio. 9. Experiment 



FRC 
85D 



An emulsion diluted in the proportion of 1 to 6 respectively with tumor serum 
and normal serimi was incubated for three hours and then inoculated into two 
series of twelve animals each and a fresh tumor emulsion into a series of 5 animals. 



INVESTIGATION OF FREUND-KAMINER REACTION 355 

Obviously the serum of normal rats exerts no deleterious 
influence upon the cells of the Flexner-Jobling carcinoma, for in 
three separate experiments cells incubated in normal serum grew 
quite as well as those incubated in serum from rats with trans- 
planted tumors. Nor was there any difference in growth between 
cells exposed to either serum and those transplanted without 
previous manipulation. 

CONCLUSIONS 

The conclusions to be drawn from the whole series of experi- 
ments are equally clear. After having made sure that there was 
no difference in the action of either rabbit or guinea-pig serum 
on the one hand, and rat serum on the other; and after having 
eliminated the sources of error enumerated by Freund and 
Kaminer, namely the admixture of connective tissue capsule and 
its blood-vessels, the presence of red blood cells in the serum, the 
use of old serum or of serum obtained at the height of digestion; 
and having, furthermore, used carcinoma (the most sensitive 
tumor to this test, according to the originators) instead of sar- 
coma, as had been done by previous investigators, we are still 
unable to detect any marked difference in the results of inocula- 
tion whether tumor serum or normal serum is used. The growth 
capacity seems neither increased in the one nor diminished in the 
other. We feel justified in stating, therefore, that the value of 
the Freund-Kaminer reaction remains at present improved. 
With this proviso, however, that these experiments prove only 
that the serum of normal rats is devoid of any deleterious effect 
on the Flexner-Jobling rat carcinoma. What might be the case 
with serum from a rat bearing a spontaneous carcinoma, we do 
not yet know. However, the first half of the experiment seems 
sufficient; for since this shows clearly that normal serum has no 
harmful effect upon the cancer cell, there can be no difference, 
with respect to cytolytic power, between normal serum and that 
of an animal with a tumor, be this transplanted or spontaneous. 

REFERENCES 

(1) Freund and Ka^miner: Biochem. Ztschr., 1910, xxvi, 312. 

(2) v. MoNAKOw: MQnchen. med. Wchnschr., 1911, Mii, 2207. 



356 LOXTIS HERLY 

(3) Stammler: Ref. MOnchen. med. Wchnschr., 1911, Iviii, 1043. 

(4) Ranzi: Handbuch d. Tech. & Meth. d. Imzn., Ergaenzungsbd., 1911, p. 611. 

(5) Kraus and v. Graff: Wien. klin. Wchnschr., 1911, xxiv, 191. 

(6) Kraus, v. Graff, and Ranzi: Wien. klin. Wchnschr., 1911, xxiv, 1003. 

(7) EIraus and Ibchiwara: Wien. klin. Wchnschr., 1912, xxy, 583. 

(8) Rosenthal: Ztschr. f. Immunitfttsforsch. u. exper. Therap., 1912, Orig., 

xiv, 174. 

(9) Ar2t and Kerl: Wien. klin. Wchnschr., 1912, xxy, 1821. 

(10) Kraus, Ischiwara, and Winternitz: Deutsch. med. Wchnschr., 1912, 

xxxviii, 303. 

(11) Coca: J. Cancer Res., 1917, ii, 61. 

(12) Frbund and Kaminer: Wien. klin. Wchnschr., 1919, xxxii, 1105. 

(13) Frankenthal: Ztschr. f. Krebsforsch., 1920, xvii, 250. 

(14) Hirschfeld: Ztschr. f. Krebsforsch., 1912, xi, 3S8. 

(15) Ischiwara: Wien. klin. Wchnschr., 1913, xxvi, 370. 

(16) KoRiTBCHoNBR AND Morqbnstern: Biochem. Ztschr., 1920, civ, 259. 

(17) Koritschoner: Biochem. Ztschr., 1922, cxxix, 605. 



IS CANCER A BIOLOGICAL PHENOMENON?^ 

SOME HERETIC THOUGHTS ON CANCER 

F. G. GADE 
Kriatiania, Secretary to the Norwegian Committee for Cancer Research 

Received for publication June 2, 1922 

The primary causes of cancer are still unknown. It seems, 
however, more and more evident that the development of cancer 
is particularly favoured by certain conditions, viz. chronic local 
irritants and old age. 

The chronic local irritanla are of the most variegated descrip- 
tions, belonging to all manifestations of energy — mechanic, 
thermic, chemical, radiological; there are also vital (animal) 
irritants (Bilharzia, Spiroptera neoplastica s. Gongylonema 
neoplasticum, Fibiger), and patholo^cal irritants (precancerous 
conditions). 

But the very difference in the nature of all these irritants indi- 
cates that it is scarcely the irritants themselves but more likely 
the reaction of the organism against these, which offers the condi- 
tions favourable for the development of cancer. Borrmann and 
Ribbert emphasize that, previous to the cancerous proliferation 
of the epithelium there is constantly foimd an inflammatory 
alteration of the underlying connective tissue, representing an 
alteration of the conditions of life of the epithelial cells. 

Very frequently, though, cancer develops in localities where 
no irritants or irritation have been noticed. 

The influence of age is more obvious; statistics from aU 
countries here agree. In Norway, the mortality from malignant 
timiors, for the years 1901-1915 is shown in table 1. 

Of these 32,274 deaths from malignant tumors, only 206, or 
0.64 per cent, have occurred in persons under thirty years of age ; 

1 Paper read in part before the Medical Society of Kristiania, September 21, 

1021. 

357 



358 



F. O. OADE 



only 1146, or 3.55 per cent, under forty years, though these groups 
include 59.27 per cent and 71 per cent respectively of the whole 
population. 

It is also remarkable that the maximum of deaths from cancer 
occurs in the period from sixty to seventy years of age, with nearly 
30 per cent (column 6) of all cancer deaths, in spite of this period of 
life numbering only 5.96 per cent (column 2) of the population. 
And in the period from seventy to eighty years there is still an 
increase as to the relative importance of cancer as the cause of 

TABLE 1 

CfUicer mortality in Norway^ lOOl-lSlS, distributed in decenni<d groups 



(1) 


(2) 


(8) 


(4) 


(S) 


(6) 


(7) 




POPUULTIOIf 
or NOBWAT 1910 


CAMCSB MOBTAUTT IN NOBWAT 


AOB 


Number of 
individuals 


Per 

hundred 

ofthe^p- 

ulation 


Total 

from 

1M1-19U 


Averace 
yearly 


Per hun- 
dred of 

allcanoer 
deaths 


Yearly 

per 
100.000 
living 


1-20 
20-30 
30-40 
40-50 
50-60 
60-70 
70-80 
80-90 
90 and more 


1, 063, 262 

334,091 

275,800 

226,181 

194,109 

140, 579 

83,825 

32,551 

3,065 

5,327 


45.10 
14.17 
11.69 
9.56 
8.24 
5.96 
3.56 
1.38 
0.13 
0.19 


38 

168 

940 

3,226 

6,925 

9,658 

8,462 

2,716 

141 


2.53 
11.20 

62.66 
215.06 
461.66 
634.85 
564.15 
181.00 
9.4 


0.12 
0.52 

2.91 

9.99 

21.45 

29.93 

26.22 

8.41 

0.45 


0.024 

0.34 

2.27 

9.55 

23.73 

45.80 

67.30 

55.60 

30.67 


No ace reported 








Total 


2, 357, 790 


99.99 


32,274 


2151 


100.00 


9.15 







death. After the age of eighty years there seems to be some 
decline in the relative cancer mortality; but this is without doubt 
only apparent, numerous deaths from cancer in the latest decen- 
niums of life being covered by the diagnosis of senile debility as 
well as by other intercurrent causes of death. 

Former investigations by the Norwegian Committee for Cancer 
Research as well as the recent ones by Dr. J0rgen Bemer make it 
very probable that the frequency of cancer continually increases 
with increasing age. 



18 CANCEB A BIOLOGICAL PHENOMENON 359 

The table shows that the ages above thirty to forty years are 
affected by cancer proportionally 226 resp. 66 times as often as 
the ages below the same years. 

Age must consequently be of the greatest importance for the 
development of cancer, and Bashford's saying, that '' cancer is a 
function of age" contains a great amount of truth. 

It must therefore be one of the nearest and most important 
objects for the cancer research work to investigate this influence 
of age. 

Surely it has been hinted that age just gives the chronic irritant 
sufficient time to act. But this explanation of the influence of 
age is hardly satisfying. We do not know the time necessary 
for any irritant to produce cancer. But suppose that even ten 
or twenty years are required, there will still be every opportunity 
for an irritant to unfold its activity in many more thousands of 
cases in the twenty to forty years group, than in the older groups, 
but cancer is very rare at the younger age. And some of 
the irritants, for example, the 2;-ray, seem to be able to pro- 
duce cancer after a proportionally short time and in younger 
individuals. 

It seems more natural to attribute the influence of age upon the 
development of cancer to the change of metabolism in the aged. 
This change alters conditions of life for cells and tissues, and a 
local irritant may add directly and indirectly to these alterations 
and accentuate the influence of senile metabolism. 

The influence of age on the development of cancer is so pro- 
noimced that it seems to signify a biological law. That cancer 
occasionally may develop in younger individuals makes no differ- 
ence as to the regular influence of age. It is not the mere number 
of years of life that constitutes ''age", but the individual's physi- 
ological condition. This senile condition may be postponed in 
many aged individuals; on the other side a senilitas praecox can 
be found in younger individuals, as we see it in the arteriosclerosis 
with young neurasthenics. 

Unfortunately we do not know enough about the metabolism 
of age to be able to form any precise conception of its influence 
upon cancer formation. ''Geriatry" has not been studied as 
thoroughly as pediatry. 



360 F. 6. GADE 

Amongst the more important somatical peculiarities of age may 
be mentioned: diminution of the total mass of blood, reduction 
of produced and needed calories, lower temperature of the body, 
greater amount of cholesterin in the fat cells, increased total 
nitrogen in the blood serum, reduction of the number of cells 
in the connective tissue and of their regenerative power, atrophy 
of the elastic tissue, of lymph nodes, spleen, Peyer's lymphatic 
noduli, and of the bone marrow. Then, also, a change and 
diminution of the fimctions of the endocrine organs. 

The senile changes appear at different times in different indi- 
viduals as well as in the different organs of the same individual. 
The senile changes can occupy the entire organism (general 
senility) or only certain organs or parts of these (local or partial 
senility). Local senility can be foimd in the female generative 
organs (ovaries, uterus, and mammse), where physiological 
activity and involution occur years before the manifestation 
of the individual's general senility. 

In accordance with this fact we see that cancer of the female 
reproductive organs appears earlier in life than cancer in the 
other organs, the great majority of deaths here occurring between 
forty and sixty years of age (Norwegian statistics), in contrast 
to deaths from cancer in other organs, where the greatest numbers 
are to be found from sixty to eighty years of age, twenty years 
later. The chorionepithelioma malignum has also its greatest 
frequency in individuals above forty years of age (Anton Sunde). 

At present we may assume that local chronic irritants, and old 
age, produce the altered conditions of cellular life which entail 
cancer. 

The real nature of these conditions is still unknown to us. But 
it may be worth remembering, that Freund and Kaminer ten 
years ago have found in the blood serum and tissue fluids of healthy 
(non-cancerous) persons a well defined fatty acid ("normal acid'Oi 
soluble in ether and having the power of dissolving cancer cells. 
This acid, that has some resemblance to the decamethylen- 
carbonic acid (COOH)s (CsHOs, of the succinic acid series, is not 
to be found in blood or tissue serum from cancerous individuals, 
whose blood has a higher alkalinity than that of non-cancerous 
individuals (Maud Menten). 



IS CANCER A BIOLOGICAL PHENOMENON 361 

We know from the experiments of Jacques Loeb and others that 
the division and multiplication of cells can be influenced by very 
small changes of their surrounding nutritive mediums. Similar 
facts are known from bacteriology. 

Robertson and Burnett have been able to accelerate the growth 
of cancerous tumors, inoculated upon rats, by injecting the animals 
subcutaneously with cholesterin (and with tethelin), substances 
which in themselves do not produce cancer in the said animals. 
Luden and Bloor have also stated that the blood of cancerous 
patients contains more cholesterin than normally. That the fat 
cells in old individuals are richer in cholesterin than those in 
younger persons has been mentioned above. 

Ferd. Blumenthal noticed that the growth of inoculated cancer 
in rats was accelerated when the animals got a surplus of potas- 
sium in their food, but retarded by a surplus of calcium. 

Direct and local influences upon the growth of the epithelium 
are known since the experiments of Fischer and others with 
Scharlach R (and other similar substances), the application of 
which on the skin of rabbits produced a lively proliferation of the 
epithelium. And Yamagiwa and Ischikawa, and now lately 
Fibiger have through similar application on rats succeeded in 
producing growth possessing the essential qualities of the genuine 
carcinoma. 

The local irritations and the old-age changes of metabolism 
entail both altered conditions of life for the cells, consequently 
also a change of their general manifestation of life. 

But such changes in general manifestations of life are just what 
we find as criteria for the transformation of a normal body cell 
into a cancer cell: changes in outer shape and structxire building 
power, changes in demand for nourishment, in speed of growth 
and proliferation, in chemical structure, enzymes and secretions 
(cachectic power), in sensibility towards electrical influences 
(x-rays), and independence of soil (metastases). 

These changes of essential biological qualities in the cancer cell 
are just what v. Hansemann has designated imder the name of 
anaplasia (transformation), a phenomenon hitherto considered as 
unique, unparalleled. 



362 F. G. OADE 

But is such a transformation really so unique in the living 
nature? 

About twenty years ago de Vries introduced into biology the 
phenomenon of mutation, the comparatively quick changes of the 
characteristics of species in living beings, in contrast to the very 
slow changes admitted by the Darwinistic principles of evolution. 
This mutation, first noticed by de Vries in a plant (Oenothera), has 
later been described for unicellular organisms, bacteria (Weisser, 
Massini), algae, spirochaetae, trepanosomas, infusoria, and 
amebas. Thj0tta and Eide have described a paratjrphoid strain 
that, under culture for about a month, essentially changed its 
character morphologically and physiologically, loosing its flagelli 
and motility and changing into a capsulated mucus-forming 
bacillus; which shape it preserved through all the consecutive 
generations. 

Gumey Dickson, in his monograph on the Transmutation of 
Bacteria, maintains that mutation is an adaption to altered 
conditions of life; it can be more or less pronounced and may 
include very different qualities of the bacteria: shape and size, 
aflSnity to dyes (chemical and physical properties of the proto- 
plasm), ability of growing on a new soil, resistance against external 
influences, virulence, and production of ferments. Most fre- 
quently mutation occurs in cultures of old date. 

There seems thus to be a considerable parallelism between mutar 
tion and the phenomena accompanying the transformations of a 
normal hody-ceU into a cancer cell. And the mutated cells, like the 
cancer cell, will transfer the newly acquired qualities to their 
future generations. 

Through all these changes in its fundamental biological 
qualities the cancer cell has turned out to quite a new cellular 
being, foreign or even hostile, to the organism in which it has 
developed. Authors like Bashford, Abderhalden, Ribbert, Butlin 
and others call the cancer cell a parasite. Butlin even described 
it as a new and special species of beings under the name of 
unicellula cancri. 

The idea that the transformation of the normal cell to a cancer 
cell may be a mutation, can certainly not explain the nature of 



IB CANCER A BIOLOGICAL PHENOMENON 363 

cancer, the nature of mutation being still enigmatical, but it 
brings the cancer question in a line with other biological phenom- 
ena, which may be examined and studied on a broader base; and 
results from other spheres of biological research may be applied 
upon the cancer questions. 

Considering the dominating influence of age on the develop- 
ment of cancer, it seems necessary for the future cancer research 
to direct the closest attention to the metabolism of the aged. 
Here will be one of the future battlefields against cancer. The 
British x-ray expert, Morley Robert, thinks it essential here to 
produce a mental and bodily youth in the individuals. The war- 
cry of the medical profession in its struggle against cancer will be: 
Keep hmnanity young I 



THE RELATION OF MUSCULAR ACTIVITY TO 

CARCINOMA^ 

A PRELIMINARY REPORT 

IVAR SIVERTSEN and A. W. DAHLSTROM 
Minneapolis, Minnesota 

Received for publication May 20, 1922 

Carcinoma in men, especially retired farmers, has been frequently 
seen by us in the past few years. These patients, men with large, 
well developed bodies of the "frontiersman" type, usually gave a 
history of having retired from the farm in good health, a few years 
previously, to enjoy a well deserved rest. On the other hand 
we have often talked with farmers, who for one reason or another 
still shouldered the responsibilities of the farm, who were well 
and working hard though sixty, seventy, or even eighty years of 
age. This observation brought us to consider muscular activity 
in relation to carcinoma. 

There have been no deaths from carcinoma among the Turn 
Teachers (1) in the past thirty years. Thirty deaths from aU 
causes occurred in this period, and we should expect at least three 
deaths from carcinoma. The Turn Teachers are of necessity 
active in a muscular sense. While the list is not large, the excep- 
tion to the rule is interesting. 

We have repeatedly noted that patients, clinically diagnosed as 
precancerous, improved in a most satisfactory manner on a pre- 
scription for increased daily exercise, as rope skipping for women^ 
golf and brisk daily walks for men. 

Muscular activity stimulates metabolism and retards the 
approach of senescence and decay. The tuberculosis problem 
has been greatly simplified by the recognition of the value of rest, 
sunshine, wholesome food and fresh air. We believe that the 

^ Read before the Hennepin County Medical Society March 6, 1922. 

366 

THB JOUBNAL Or CAKCBB BBBXABCH, TOL. TI, MO. 4 



366 IVAR SIVERTBEN AND A. W. DAHL8TROM 

carcinoma problem may also be simplified by the recognition 
of the value of muscular activity in those periods of Kfe most 
affected by carcinoma, those periods of life when we are prone to 
*'let down*' and check our pace. If muscular activity has an 
intimate relationship to carcinoma, the value of it will be most 
apparent in prophylaxsis, the field which to date promises the 
most in the attack on this malady. 

In a perusal of the literature we are unable to find any direct 
reference to the relation of muscular activity to carcinoma.* 
Hoffman (2) shows that carcinoma is world wide, is a disease of 
maturity, affecting both sexes and seriously affecting all civilized 
nations and races. Observations are quoted and statistics are 
presented showing that carcinoma is more frequent among the 
unoccupied, among people of independent means, among the 
leisured and professional classes, among the ovemourished, weU 
housed, well fed, and among those who live in the warmer climates. 
Carcinoma is less frequent among the people who are actively 
engaged in a gainful occupation, among the hard working classes, 
among those who live in large industrial centers, among peoples 
predominatingly poor and of necessity frugal, living on an alimen- 
tation just sufficient for their actual needs, and among those who 
live in the colder climates where the struggle for existence is more 
difficult. Reading between the lines one common factor is noted, 
that of the varjdng degree of muscular activity. The vast 
majority of people follow the line of least resistance and those of 
independent means, the leisured and professional classes, and 
those who live in the tropics are not compelled to do manual labor 
as a means of gaining a living and hence are less liable to be active 
in a muscular sense. Attention must be called to Dublin's (3) 
conclusion with reference to the experience of the Metropolitan 
Life Insurance Company for 1914, 1915, 1916, that "the cancer 

* It ia interesting to us to find that in another part of the world aimilar ideas 
have been advanced almost simultaneously. Cherry, of Melboumei on March 1, 
1922, read a paper on A Theory of Cancer (published in Med. Jour, of Australia, 
April 22, abstracted in J. A. M. A., 1922, Ixxix, 245) in which he attributes 
the recent increase of cancer to deficient muscular activity, ovemutrition, 
and lessened skin activity, as a result of modem living conditions of civilised 
peoples. 



REIATION OF MUSCXTLAR ACTIVITr TO CARCINOB£A 367 

mortality rate at ages where the cancer rate is significanti 
decreases as we go up the economic scale." To reconcile the 
statistics presented by Hofifman with those of Dublin one must 
admit that other factors than the economic status are present. 

Ewing (4) states that ''statistics show that sanitary measures 
which control infectious diseases exercise no such power over 
carcinoma, which is somewhat conspicuous by a relative failure to 
attack the poor, the overworked, the underfed, the savage, but 
chooses a notable proportion of its victims among the well-to-dQ| 
the well nourished, the well protected against infectious di9eases, 
and the indolent." 

Ewing has noted also that carcinoma seemis to be as widespread 
aa are animal species. The mortality from malignant timxors 
among pet dogs is r^orted to be surprisingly high. Forms of 
carcinoma in horses, cattle, and swine are relatively common, 
carcinoma of nearly every organ of fowls has been reported, mice 
and rats are subject to carcinoma and made use of extensively 
in the laboratory. It is very interesting to note that he too calls 
attention to the rather sharp faU in carcinoma incidence when 
one takes up the consideration of carcinoma in wild animals. 
In this respect one must also consider that wild animals have not 
been subjected to uniform close study, autopsies have been less 
frequent, and violent deaths have been more common than among 
the domesticated animals. McCoy (5) reports autopsy findings 
of 100,000 rats (95 per cent of which were Mus norvegicus) caught 
during the San Francisco rat campaign about fifteen years ago. 
Of these rats 103 were found to have tumors, of which 10 proved 
to be carcinoma. Thus he found that about one rat for every 
thousand had a new growth, and only 1 in 10,000 had a carcinoma, 
the remainder being benign tumors or sarcomata. In addition 
to the factors incident to the study of carcinonoa among wild 
animals, the necessary muscular activity of the animal in its 
natural wild state may have a bearing on the relative inf requency 
of reported carcinoma. 

Carcinoma has been observed under certain conditions among 
artificially bred fish (6). It has been clearly shown that fish 
kept under artificial conditions are more susceptible to the 



^368 IVAK SIVERTSEN AND A. W. DAHLSTROM 

development of carcinoma than are those livmg under natural 
conditions. We believe that the enforced inactivity of the fish 
in ponds and hatcheries is a potent factor in the development of 
carcinoma. The dietery factor must also be considered. 

The present day labor saving machinery, convenient means of 
transportation, and ultra-conveniences of modem life all tend to 
make us less active physically, and it appears to us that the 
increasing curve of carcinoma closely approximates the inaugura- 
tion of the Age of Machinery. 

The recent researches into the finer biochemical changes 
attending the metabolism of carcinomatous patients point to the 
development of carcinoma as an individual biological reaction to 
chronic irritation, governed possibly by the previous state of 
metabolism. It would seem that we must look to the biochemist 
for information as to the exact cause of the reversion of the adult 
cell to fetal characteristics. The researches of Freimd and 
Kaminer (7), Rosenthal (8), Neuberg (9), Koritschoner and 
Morgenstem (10), and others are very interesting in this 
connection. 

This opportunity is taken for the presentation of a working 
hyi>othesis: That human carcinoma ma/ be the reaction to and 
the result of chronic irritation of adult epithelial tissues bathed in 
body fluids altered by certain metabolic products as a result 
of deficient muscular activity. 

In advancing the above hypothesis constructive and destructive 
criticism is invited, as by intensive cooperative effort sufiKcient 
data pro and con will be obtained definitely to prove or disprove 
our contention. 

An investigation in an experimental way of some of the phases 
of muscular activity and its relation to carcinoma is being con- 
ducted which will be reported in the near future. We refer 
especially to the investigation of the effect of muscular activity 
on experimental carcinoma in mice. We are also preparing to 
ascertain the effect on human carcinoma of horse serum obtained 
after graduated amounts of muscular activity. 

In an effort at this time to gain some concrete evidence bearing 
on the subject, a study of carcinoma in Minnesota for the past 



RELATION OF MUSCXJLAB ACTLYITT TO CABCINOBCA 369 

three yeaxs was undertaken. Access to original certificates of 
death was obtained through the courtesy of the Bureau of Vital 
Statistics, Minnesota State Board of Health, there being no 
compilations available. Among 86,838 records examined 6,351 
deaths were ascribed to malignant disease for the years 1918, 
1919, and 1920. Sarcoma was named as cause of death in 452 
records. There were 3135 deaths from carcinoma among males, 
and these were used as the basis for the present study. Car- 
ciDoma among females was not studied at this time because of the 
difficulty in estimating the amount of muscular activity rela- 
tive to occupations of this sex. The deaths were tabulated 
according to occupation and the occupations divided by us into 
six groups according to our idea of the muscular activity 
necessary to that occupation. It is to be regretted that no 
scientific criterion was available for this classification. Some 
method as suggested by Waller and DeDecker (11) might be 
applicable if standardized by sufficient usage. We have tried to 
use our best judgment in each case. A special report was 
obtained from the Bureau of Census on the number of persons 
engaged in each occupation based on the Census of 1920. The 
deaths in each group for three years were totalled, averaged, the 
number of persons engaged calculated and the death rate com- 
puted on the basis of 100,000 persons engaged to briQg each group 
to a common ground for comparison. The death rates are id a 
measure not absolutely reliable because of the short period of 
time imder consideration and the present day laxness of the 
average undertaker in completely filling out certificates of death. 
We grouped the occupations as follows: Group 1. Occupations 
involving great muscular activity. In this group are included 
stonecutters, blacksmiths, boilermakers, moulders and the Uke. 
Group 2. Occupations involving moderate amounts of muscular 
activity, as carpenters, cabinet makers, common laborers, masons, 
pliunbers, and the like. Group 3. Includes occupations involv- 
ing medium amounts of muscular activity as foremen, millers, 
brewers, agents, printers, etc. Group 4. Occupations involving 
small amounts of muscular activity, as the professions, office 
workers, and the like. Group 5. The farmers, because of the 



370 IVAtt SIVERTSEN AND A. W. DAHLSTROM 

seasonal character of their work in Minnesota, were classed alone. 
Group 6. This group includes those not actively engaged in any 
gainful occupation, students, idiots, epileptics, inmates of insane 
hospitals, and those who have retired from gainful occupations. 
Hoffman (12) reports the following statistics: 

Death raU per 100,000 males, Bngland'Walee, ihiriy-fioe years of age and over 




All males 165.5 210.3 

Occupied 146.0 180.0 

Not occupied 859.0 503.8 

----■'■ - - - - 

The differences in mortality are very striking and we believe 
they may be explained on tiie basis of muscular activity. 

In Minnesota for the years 1918, ldl9, 1920 we have found the 
death rate per 100,000 males as follows: 

1. Those 21 years of age and over: 

Active (groups 1, 2, 3, 4, 5) 103.5 

b. Inactive (group 6) 510.0 

2. Those 45 years of age and over: 

a. Active (groups 1, 2, 3, 4, 5) 285.8 

b. Inactive (group 6) 681 .0 

The figures for Minnesota are seen to parallel roughly those for 
England and Wales (see chart 1). 

In a consideration of the age class 21 years of age and over we 
have found the death rate from carcinoma lowest in group 1 
(in which the necessary muscular activity is greatest) and bluest 
in group 6 (in which the necessary muscular activity is least). 
There is a stairstep gradation between the extremes. The ratios 
between the death rates in the various groups may be thus 
expressed; I:II:III:IV:V: VI: 1::1: 1.7: 2.1:2.3:3.6: 11.8 (see 
chart 2, table 1). 

To ascertain the ratios in the various i^e classes we have 
subdivided the active groups: 

A. Those twenty-one to forty-four years of age. In this age 
class the ratios may be expressed thus: I:II:III:IV:V: :1:1.4: 
1.5:2.2:2.0 (see table 2, chart 3). 



RELATION OF MUSCULAR ACTrVITT TO CABCINOICA 



371 



DEATH RATES PER 100,000 MALES 
Bngland-Wales Minnesota 
1890*1892 1900-1902 1918-1919-1920 


£00. 






































B/S, 














1 




































































































H 






































H 




























1 


n 








H 


















B 












1 








'1 ' 






































^■__ 


490 






























1 








^■'"^ 


































H 


































H 


































H 


590 








_g_ 


























H 


































H 
































A 


H 






























2 


H 


fiOO 
100 
































■ 






























H 


















. 






















A 




















A 












































^ 








_pg 












1 










■-■ 




- 






























BJ 






-■-l 
































M M 

















Chart 1. Columns lettered A denote occupied; B, not occupied; columns not 
numbered denote those thirty-five years of age and over; numbered J, twenty- 
one years of age and over; i, forty-five years of age and over. 



-372 



IVAR SIVERT8EN AND A. W. DAHIiSTROH 



BIQ 




















"^*" 


lyj 


iSfl 




















— 


■E 




OBATH RATB8 PflR 
100,000 M/XB3, 
ei TBARS or AGS 
AHD OVBR* MIBN. 
1918-1919-1980. 


— 


;f 


SOQ, 


- 

-1 




"~B~ 




H^ 




""^r 




-W- 










































"Bz 






















""'^w^ 






















■ B~ 






















B^ 






















LB!!I 


ISfi 






















^^M 






















^^^1 






















^■^ 






















""■^ 






















H~~ 






















' "H" 






















'""Bl 






















■~ 




















y_ 


"T 








































im 




















til 












11 




IV 




rr 








11 






■i 


■~H~] 


w 












■ H~H~Hn 




J 




■ ■■■■' 




— 


1 




1 1 1 1 B^ 

1 1 1 mm 



Chart 2. CoLuifNs Numbered I, II, etc., Refer to Occupation Groups 1, 2, etc. 



TABLE 1 

Death rates per 100,000 mdles, twenty-one years of age and over, Minnesota, 1918, 

1919, 19$0 



OBODF 


NUM BBB OP PEBSONB 


DBATHB 


DIATB BAT! 


1 


46,736 


20.0 


43.0 


2 


168,460 


128.0 


75.8 


3 


84,523 


77.0 


91.1 


4 


211 ,824 


213.9 


100.8 


5 


170,483 


267.7 


156.6 


1.2,3,4,5 


682,026 


706.4 


103.5 


6 


64,503 


330.0 


510.0 



REIATION OF MUSCULAR ACTIVITT TO CARCINOMA 373 

B. Those forty-five to sixty-four years of age. The ratios 
existmg are I:II:III:IV:V::1:1.8:2.1:2.0:2.0 (see chart 4, 
table 3). 

C. In those sixty-five years of age and over we have found the 
difiference less striking as might be expected for this age class, 
the ratios being: I:II:III:IV:V::1:1.5:1.1:1.4:1.9 (see table 
4, chart 5). 

D» Considering those of another age class, from twenty-one 
to sixty-four years of age, we find there is progressive increase in 
death rate with progressive diminution in amount of necessary 
muscular activity. The ratios may be expressed: I: II: III: IV: 
V: : 1: 1.6:2.3:2.4:3.1 (see table 5, chart 6). 

In all the age classes considered the death rate from carcinoma 
among those actively engaged in a gainful occupation is roughly 
inversely proportional to the amount of muscular activity 
necessary to that occupation. The only out and out exception to 
the rule is in group 5, the farmers, for which we lack a satisfactory 
explanation except that the seasonal character of the occupation 
in Minnesota may have a bearing, in that during the spring and 
fall seasons the work is very heavy, while in the winter and sum- 
mer, especially in winter, the work is very Ught. 

The experience of the Industrial Department of the Metro- 
politan life Insurance Company (13) discloses that cancer causes 
the following percentage of total deaths for the following occu- 
pations: 

Group 1. Blacksmiths 7.6 per cent; miners 4.6; moulders 3.9; 
average for the group 5.9 per cent. 

Group 2. Laborers 5.5; longshoremen 4.3; machinists 4.5; 
masons 6.5; average 5.2 per cent. 

Group 3. Teamsters 3.6; printers 2.7; painters 4.1; track 
laborers 4.7; average 3.8 per cent. 

Group 4. Clerks 3.1 ; cigarmakers 5.2; average 4.2 per cent. 

Group 5. Fanners 7.6 per cent. 

The above figures do not indicate any dkect influence of the 
occupational disease factor. 



Cbart 5 Chart S 

Chart 3. Goluunb Nuiibxred I, II, kic., Refer to Groups 1, 3, etc. 
Chart 4. Columns Nvuberep I, II, etc.. Refer to Groups 1, 2, etc. 
Chart 5. Columns Numbered I, II, etc., Refer to Groups 1, 2, etc. 
Chart 0. Columns Numbered I, II, etc., Refer to Groups 1, 2, ivc. 
374 



RELATION OP MITSCXJIAR ACTIVITY TO CARCINOMA 



375 



TABLE a 

Deaih raUa per 100,000 nudes engaged, twenty-one to forty-four years of age, Minne* 

sota 1918, 1919, 19tO 



ttftOtIF 


vniBBB or pxasom 


DBATBB 


DBATRBATB 


1 


34,587 


2.7 


7.8 


2 


128,233 


14.3 


11.1 


3 


58,362 


6.7 


11.5 


4 


141,902 


24.7 


17.4 


5 


93,661 


16.7 


16.7 



TABLES 

Deaik rales per 100,000 males engaged, forty-Jwe to sixty-four years of age, MxnnS' 

sota, 1918, 1919, 19i0 



eaOKfpB 


IfXTMBUI OP PKBBOm 


DAma 


DBATBBATB 


1 


11,057 


11.3 


102.2 


2 


4,373 


66.3 


189.0 


3 


22,055 


61.0 


221.1 


4 


61,321 


126.0 


205.5 


5 


66,168 


136.0 


205.5 



TABLE 4 

DeaA tales per 100,000 males engaged, sixty-fioe years of age and over, Minnesota, 

1918, 1919, 1990 



QBOUn 


aruMBBB or PBaaom 


DBATBS 


DBA9VBATB 


1 


1,092 


6.0 


549.4 


2 


5,854 


48.3 


825.0 


3 


3,206 


19.3 


602.0 


4 


8,601 


65.0 


755.7 


5 


10,654 


115.7 


1065.0 



TABLES 

Death, rates per 100,000 males engaged, twenty-one to sixty-four years of age, Minne^ 

sota, 1918, 1919, 19t0 



QBOUn 


MUIIBBB or PXRBOKB 


DBATHB 


DBATBBAm 


1 


45,644 


14.0 


30.6 


2 


162,606 


79.7 


49.0 


3 


81,317 


57.7 


70.9 


4 


203,223 


148.7 


73.1 


5 


159,829 


151.7 


94.9 



.376 



IVAR SIVERTSEN AND A. W. DAHLSTROM 



A tabulation of the ages at death in the different occupation 
groups discloses a rather close approximation of extremes and 
averages so it appears the differences in death rates shown above 
can not be wholly explained on a basis of age incidence; that is, 
there is not sufficient difference between the ages of those in the 
different occupation groups to explain the ratios found to exist 
between the death rates in the various groups of occupations 
(see table 6). 

TABLES 

Ages at death, males, Minnesota, 1918, 1919, 1990 



QBOUPS 


XniCBBB or PIATKS 


■xrBBma 


▲YBBAOB AQB AT DBATH 


1 


67 


24-81 


67.3 


2 


879 


29-01 


60.3 


3 


201 


29^ 


67.7 


4 


607 


24-88 


66.8 


6 


878 


2&-M 


62.2 


6 


084 


27-103 


70.4 



SUMMARY 

In a preliminary report attention is called to a new factor in 
carcinoma etiology. Carcinoma has been frequently observed in 
retired farmers, seldom in active individuals. Precancerous 
patients have improved clinically with increased muscular 
activity. Attention is called to observations of less carcinoma 
among those of necessity physically active. The reported 
incidence among domesticated animals is greater than in wild 
animals. A working hypothesis is advanced: That human car- 
cinoma may be the reaction to and the result of chronic irritation 
of adult epithelial tissue bathed in body fluids altered by certain 
metabolic products as a result of deficient muscular activity. 
From a study of carcinoma deaths among males in Minnesota for 
three years it appears that the death rate in those who are active 
is greatly exceeded by the death rate in those who are in- 
active. From a study of the death rates of those who are 
actively engaged in a gainful occupation it appears that the death 
rate is lowest in those occupations involving the greatest amounts 
of necessary muscular activity, and is highest in those occupations 
involving the least amounts of muscular activity. The age 



RELA'nON OP MUSCtJLAR ACTIVITY TO CABCINOMA 877 

incidence factor of the cases studied does not explain the varia- 
tions shown. The figures from the experience of the Metropoli- 
tan life Insurance Company do not show any marked influence 
of occupational disease on carcinoma death rate. Additional 
study and accumulation of data are necessary to establish 
definitely the status of the relation of muscular activity to 
carcinoma. 

CONCLUSIONS 

1. Carcinoma constitutes a serious menace to the adults of 
all civilized races. 

2. It appears that the recent increase has accompanied the 
advent of the Age of Machinery. 

8. The reported incidence among the lower animals appears 
inversely proportional to the degree of muscular activity necessary 
to the existence of the animal. 

4. The death rate in males actively engaged in a gainful 
occupation is less than the death rate in those not actively engaged 
in any gainful occupation. 

« 

5. The death rate among males actively engaged in a gainfid 
occupation is inversely proportional to the degree of muscular 
activity necessary for that occupation. 

6. A new factor in carcinoma etiology is proposed. 

7. A working hjrpothesis is suggested. 

REFERENCES 

(1) Personal Communication from the Secretary of the Turn Teachers Mutual 

Benefit Association. 

(2) HorFMAN, F. L.: The Mortality from Cancer Throughout the World. 

Newark, 1915. 

(3) DiTBLiN, L. I.: Mortality Statistics of Insured Wage Earners and their 

Families. Metropolitan Life Insurance Company, 1910, Chapter iz. 

(4) EwzNo, J.: Animial Experimentation and Cancer. Defense of Research, 

Pamphlet iv, A. M. A., 1911, pp. 4-6. 

(5) McCoT, G. W.: A preliminary report on tumors found in wild rats. Jour. 

Med. Res., 1909, xvi, 285. 

(6) Bainbridox, W. S.: The Cancer Problem. New York, 1914, p. 55. 

(7) Fbxxjnd, E., and Kamineb, G. : Reactions between tumor cells and blood 

serum. Biochem. Ztschr., 1910, xxvi, 312. 

(8) Rosenthal, E.: The biological parallel between cancer and fetal cells. 

Ztschr. f. Immunit&tsforsch. u. exper. Therap., 1912, xiy, 174. 



378 IVAB SIVSBTBEN AND A. W. DAHL8TROM 

(0) Neubbro, C: Chemistry of tumon. Bioohem. ZtBchr., 1010, xxyi, 344. 

(10) KoBiTscHONBB, I. B., AMi> MoBGBNBTBBM, O.: Befractometiic studies of the 

Freund-Kaminer reaction. Biochem. Ztschr., 1920, eiv, 259. 

(11) Walleb, A. D., AND DbDbcxeb, G.: The physiologicai cost of muscular 

work. Brit. Med. Jour., 1921, i, 669; 1921, ii, 627. 

(12) Hoffman, F. L.: The mortality from cancer in different occupations. 

Ztschr. d. osterr. Sanitatswesen, 1918, supplement to 1-26, pp. 33-42. 

(13) Dublin, L. I. : Causes of death by occupation. Bull. U. S. Bureau Labor 

Statistics, No. 207, U. S. Government Printing OflSce, 1917. 



INDEX 

Actiyity, muscular, The relation of, to carcinoma. A preliminary report 305 

Adenoma, atypical. An, of the pancreas originating in islet tissue 277 

Adrenal, Primary spontaneous tumors in the kidney and of mice 305 

Age, Teratomas and their relation to 261 

Albino rats, The influence of certain diets upon tumor susceptibility and 

grovrth in (summary) 100 

Animal tumors,- transplanted. The effect of combined heat and radiation 

upon (summary) 101 

Animals, immune, Effect of blood from, upon transplantable tumors 25 

Bagg, Halsey J., Interstitial injections of an actiye deposit of radium emana- 
tion in a rat carcinoma (summary) 104 

Benedict, Stanley R., and Sugiura, K., The influence of certain diets upon 

tumor susceptibility and growth in albino rats (summary) 100 

Bemhard, A., Rohdenburg, G. L., and Krehbiel, O. F., Further investigations 
of disturbances of blood sugar equilibrium in their relation to neoplasia. . 223 

Biological phenomenon. Is cancer a. Some heretic thoughts on cancer 357 

Biology, tumor, A phase of (summary) 00 

Bladder, cancer of the, Radium in (summary) 100 

Blood, Effect of, from immune animals upon transplantable tumors 25 

sugar equilibrium, Further investigations of disturbances of, in their 

relation to neoplasia 223 

y The protein content of the whole, and plasma in cancer 127 

Bovie, William T., Further experiments on sensitisation to heat by radiation 

(summary) » 103 

Breast, cancer of the, A few late results after the radical operation 

for (summary) 88 

Bulkley, L. Duncan, Proofs of the constitutional nature of cancer (summary) 185 

Burrows, Montrose T., Problems in cancer research 101 

Burrows, Montrose T., Problems in cancer research 131 

Cancer and parasite 257 

, Is, a biological phenomenon? Some heretic thoughts on cancer 357 

-^-, mortality. Is, increasing? 251 

— ', of the bladder, Radium in (summary) 190 

, of the breast, A few late results after the radical operation for (summary) 88 

, operations, End-result reports of (summary) 187 

^ Proofs of the constitutional nature of (summary) 185 

, research. Problems in 101 

, research, Problems in 131 

, Research, Proceedings of the American Association for. Fourteenth 

Annual Meeting 175 

379 



' 



380 INDEX 

Cancer, Research, Proceedings of the American Association for, Thirteenth 
Annual Meeting 87 

, The protein content of the whole blood and plasma in 127 

Carcinoma, A comparison of the growth of mice which ultimately develop, 
with the growth of mice which do not develop carcinoma 7 

, Primary spontaneous squamous cell, in mice. Studies on the incidence 

and inheritability of spontaneous tumors in mice , 57 

, rat, Interstitial injections of an active deposit of radium emanation 

in a (summary) 104 

, squamous-cell, The occurrence of, in mice (summary) 180 

, The influence upon the growth of transplanted Flexner-Jobling rat, of 

hydrogen ions and of various salts in different concentrations 285 

-, The relation of muscular activity to. A preliminary report 305 



Cell, fat, A study of a lipomyxosarcoma with conmients upon the origin of 
the 109 

Chromatophores, Inoculation of sarcomatous tumors into negro fowls, with 
special reference to the significance of 31 

Dahlstrom, A. W., and Sivertsen, Ivar, The relation of muscular activity to 
carcinoma. A preliminary report 305 

Diets, The influence of certain, upon tumor susceptibility and growth in 
albino rats (summary) 100 

Duane, William, High frequency x-ray spectra 170 

Falk, K. George, Sug^iura, Kanematsu, and Noyes, Helen Miller, The influence 
upon the growth of transplanted Flexner-Jobling rat carcinoma of 
hydrogen ions and of various salts in different concentrations 285 

Fat cell, A study of a lipomyxosarcoma with comments upon the origin of 
the 109 

Flexner-Jobling rat carcinoma, The influence upon the growth of trans- 
planted, of hydrogen ions and of various salts in different concentrations. 285 

Fowls, negro, Inoculation of sarcomatous tumors into, with special reference 
to the significance of chromatophores 31 

Freund-Kaminer reaction, A critical investigation of the 337 

Fujii, Tamotsu, andKon, Yutaka, Inoculation of sarcomatous tumors into 
negro f owb, with special reference to the significance of chromatophores . 31 

Gade, F. G., Is cancer a biological phenomenon? Some heretic thoughts 

on cancer 357 

Goldblatt, Harry, An atypical adenoma of the pancreas originating in islet 

tissue 277 

Greenough, Robert B., and Simmons, Channing C, End-result reports of 

cancer operations (summary) 187 

Growth of mice, A comparison of the, which ultimately develop carcinoma 

with the growth of mice which do not develop carcinoma 7 

, of transplanted Flexner-Jobling rat carcinoma, The influence upon the, 

of hydrogen ions and of various salts in different concentrations 285 

f rate of transplanted spontaneous tumors in mice, Effect of a reduction 

of lymphocytes on 1 



INDEX 381 

Heat and radiation, combined, The effect of, upon transplanted animal 

tumors (summary) 101 

, sensitization to, Further experiments on, by radiation (summary) 193 

Heredity, The influence of, in determining tumor metastasis 139 

, The influence of, in determining tumor metastasis (simunary) 182 

Herly, Louis, A critical investigation of the Freund-Kaminer reaction 337 

Heterosexual characters, Relation of, to neoplasms and developmental 

errors (summary) 186 

Himwich, H. £., Teratomas and their relation to age 261 

Holmes, Harriet F., Slye, Maud, and Wells, H. Gideon, Primary spontaneous 
squamous cell carcinomas in mice. Studies on the incidence and in- 
heritability of spontaneous tumors in mice. Fifteenth communication.. 57 

, Slye, Maud, and Wells, H. Gideon, Primary spontaneous tumors in the 

kidney and adrenal of mice. Studies on the incidence and inheritability 

of spontaneous tiunors in mice. Seventeenth communication 305 

, Slye, Maud, and Wells, H. Gideon. Primary spontaneous tumors of the 

ovary in mice. Studies in the incidence and inheritability of spontaneous 

tumors in mice. Fourteenth report (summary) 02 

, Slye, Maud, and Wells, H. Gideon, The occurrence of squamous-cell 

carcinoma in mice (simunary) 180 

Huffman, Otto V., Relation of heterosexual characters to neoplasms and 

developmental errors (summary) 186 

Hydrogen ions. The influence upon the growth of transplanted Flexner* 
Jobling rat carcinoma of, and of various salts in different concentrations. 285 

Inoculation of sarcomatous tumors into negro fowls, with special reference to 
the significance of chromatophores 31 

Ions, hydrogen. The influence upon the growth of transplanted Flexner-Job- 
ling rat carcinoma of, and of various salts in different concentrations 285 

Islet tissue. An atypical adenoma of the pancreas originating in 277 

Jacobson, Victor C., A study of a lipomyxosarcoma with comments upon the 
origin of the fat cell 109 

Kellert, Ellis, Influence of the lymphocyte on the peritoneal implantation 

of sarcoma in mice 41 

Kidney and adrenal of mice, Primary spontaneous tumors in the 305 

Knox, L. C., Massage and metastasis (summary) 192 

Kon, Yutaka, and Fujii, Tamotsu, Inoculation of sarcomatous tumors into 

negro fowls, with special reference to the significance of chromatophores . 31 
Krehbiel, O. F., Rohdenburg, G. L., and Bemhard, A., Further investigations 
of disturbances of blood sugar equilibrium in their relation to neoplasia. . 223 

Kross, Isidor, Cancer and parasite 257 

, Effect of blood from immune animals upon transplantable timiors 25 

, Parabiosis and tumor growth 121 

, Pregnancy and tumor growth 245 

Levin, Isaac and Levine, Michael, The r61e of neoplasia in parasitic diseases 
of plants 95 



382 INDEX 

Leyine, Michael, and Levin, Isaac, The r61e of neoplasia in parasitic 

diseases of plants 95 

Lipoxnyxosarcoma, A study of a, with comments upon the origin of the fat 

cell 109 

Little, C. C, Factors underlying susceptibility to a transplantable tumor in 

mice (summary) 106 

Loeb, Leo, Further inyestigations on the origin of tumors in mice. VII* 

Tumor age and tumor incidence 197 

Lymphocyte, Influence of the, on the peritoneal implantation of sarcoma 

in mice 41 

Lymphocytes, Efifect of a reduction of, on the growth rate of transplanted 

spontaneous tumors in mice 1 

Massage and metastasis (sxunmary) 192 

Metastases, tumor, The influence of heredity in determining 1S9 

Metastasis, Massage and (sununary) 192 

, tmnor, The influence of heredity in determining (summary) 182 

Meyer, Willy, A few late results after the radical operation for cancer of the 

breast (summary) 88 

Mice, A comparison of the growth of, which ultunately develop carcinoma 

with the growth of mice which do not develop carcinoma 7 

, Effect of a reduction of lymphocytes on the growth rate of transplanted 

spontaneous tumors in 1 

, Factors underlying susceptibility to a transplantable tumor in 

(summary) 106 

, Further investigations on the origin of tumors in 197 

, Influence of the lymphocyte on the peritoneal implantation of sarcoma 

in 41 

, Primary spontaneous squamous cell carcinomas in. Studies on the inci- 
dence and inheritability of spontaneous tumors in mice 57 

, Primary spontaneous tumors in the kidney and adrenal of 305 

, Primary spontaneous tumors of the ovary in (summary) 92 

, Studies in the incidence and inheritability of spontaneous tumors in 139 

, Studies in the incidence and inheritability of spontaneous tumors in 

(summary) 182 

, The occurrence of squamous-cqf 1 carcinoma in (summary) 180 

Muscular activity, The relation of, to carcinoma. A preliminary report . . . 365 

Negro fowls. Inoculation of sarcomatous tumors into, with special reference 
to the significance of chromatophores 31 

Neoplasia, Further investigations of disturbances of blood sugar equilibrium 
in their relation to 223 

, The r61e of, in parasitic diseases of plants 95 

Neoplasms, Relations of heterosexual characters to, and developmental 
errors (sununary) 186 

Noyes, Helen Miller, Sugiura, Kanematsu, and Falk, K. George, The influence 
upon the growth of transplanted Flexner-Jobling rat carcinoma of hydro- 
gen ions and of various salts in different concentrations 285 



INDEX 383 

Operations, cancer, End-reeult reports of (summary) 187 

Orary, Primary spontaneous tumors of the, in mice (summary) 92 

Pancreas, An atypical adenoma of the, originating in islet tissue 277 

Parabiosis and tumor growth 121 

Parasite, Cancer and 257 

Parasitic diseases of plants, The r61e of neoplasia in 05 

Peritoneal implantation of sarcoma in mice, Influence of the lymphocyte 

on the 41 

Phenomenon, biological, Is cancer a. Some heretic thoughts on cancer 357 

Plants, The r61e of neoplasia in parasitic diseases of 95 

Plasma, whole blood and. The protein content of the, in cancer 127 

Pregnancy and tumor growth 245 

Prime, Frederick, A phase of tumor biology (summary) 96 

, Effect of a reduction of lymphocytes on the growth rate of transplanted 

spontaneous, tumors in mice 1 

, and Rohdenburg, George L. The effect of combined heat and radiation 

upon transplanted animal tumor (summary) 101 

, and Wood, F. C, Further studies in radiation dosage (summary) 177 

Proceedings of the American Association for Cancer Research, Thirteenth 

Annual Meeting 87 

Report of the council 87 

A few late results after the radical operation for cancer of the breast 88 

Primary spontaneous tumors of the ovary in mice. Studies in the incidence 
and inheritability of spontaneous tumors in mice. Fourteenth report ... 92 

The r61e of neoplasia in parasitic diseases of plants 95 

A phase of tumor biology 96 

The influence of certain diets upon tumor susceptibility and growth in 

albino rats 100 

The effect of combined heat and radiation upon transplanted animal 

tumors 101 

Interstitial injections of an active deposit of radium emanation in a rat 

carcinoma 104 

Factors underlying susc€fi>tibility to a transplantable tumor in mice 106 

Proceedings of Uie American Association for Cancer Research, Fourteenth 

Annual Meeting 175 

Report of the council 175 

Report on the Journal or Cancer Rbbbarch 177 

Further studies in radiation dosage 177 

High frequency x-ray spectra 179 

The occurrence of squamous-cell carcinoma in mice 180 

The influence of heredity in determining tumor metastasis. Studies in 
the incidence and inheritability of spontaneous tumors in mice. Fif- 
teenth report 182 

Proofs of the constitutional nature of cancer 185 

Relation of heterosexual characters to neoplasms and developmental 

errors 186 

End-result reports of cancer operations 187 

Radium in cancer of the bladder 190 



I 



384 INDEX 

Proceedings of Fourteenth Annual Meeting — continued 

Problems in cancer research 191 

Massage and metastasis 192 

Further experiments on sensitization to heat by radiation 193 

Malignant timiors of the thyroid 194 

Experimental production of timiors 196 

Protein content, The, of the whole blood and plasma in cancer 127 

Radiation, combined heat and, The effect of, upon transplanted animal 
tumors (summary) 101 

, dosage, Further studies in (summary) 177 

, Further experiments on sensitization to heat by (summary) 193 

Radium emanation. Interstitial injections of an active deposit of, in a rat 
carcinoma (summary) 104 

f in cancer of the bladder (summary) 190 

Rat carcinoma, Flexner-Jobling, The influence upon the growth of trans- 
planted, of hydrogen ions and of various salts in different concentrations . 285 

, carcinoma. Interstitial injections of an active deposit of radium emana- 
tion in a (summary) 104 

Rats, albino, The influence of certain diets upon tumor susceptibility and 
growth in (summary) 100 

Ray, L. A., and Robertson, T. Braibford., A comparison of the growth of 
mice which ultimately develop carcinoma with the growth of mice which 
do not develop carcinoma 8 

Reaction, Freund-Kaminer, A critical investigation of the 337 

Robertson, T. Brailsf ord, and Ray, L. A., A comparison of the growth of mice 
which ultimately develop carcinoma with the growth of mice which do 
not develop carcinoma 7 

Rohdenburg, George L., and Prime, Frederick. The effect of combined heat 
and radiation upon transplanted animal tumors (summary) 101 

Rohdenburg, G. L., Krehbiel, O. F., and Bemhard, A., Further investiga- 
tions of disturbances of* blood sugar equilibrium in their relation to 
neoplasia 223 

Salts, The influence upon the growth of transplanted Flexner-Jobling rat 

carcinoma of hydrogen ions and of, in various concentrations 285 

Sarcoma, Influence of the lymphocyte on the peritoneal implantation of, in 

mice 41 

Sarcomatous tiunors, Inoculation of, into negro fowb, with special reference 

to the significance of chromatophores 31 

Sensitization to heat by radiation. Further experiments on (summary) 193 

Simmons, Channing C, and Greenough, Robert B., End-result reports of 

cancer operations (summary) 187 

Sivertsen, Ivar, and Dahbtrom, A. W., The relation of muscular activity to 

carcinoma. A preliminary report 365 

Slye, Maud, The influence of heredity in determining tumor metastases. 

Studies in the incidence and inheritability of spontaneous tumors in mice. 

Sixteenth report 139 



INDEX 385 

Slye, Maud, The influence of heredity in detennining tumor metastasis. 
Studies in the incidence and inheritability of spontaneous tumors in 
mice. Fifteenth report (summary) 182 

, Hohnes, Harriet F., and Wells, H. Gideon. Primary spontaneous 

squamous cell carcinomas in mice. Studies on the incidence and in- 
heritability of spontaneous tumors in mice. Fifteenth communication.. 57 

, Holmes, Harriet F., and Wells, H. Gideon, Primary spontaneous tumors ^ 

in the kidney and adrenal of mice. Studies on the incidence and in- 
heritability of spontaneous tumors in mice. Seventeenth coiomunication 805 

, Holmes, Harriet F., and Wells, H. Gideon. Primary spontaneous 

tumors of the ovary in mice. Studies in the incidence and inheritability 
of spontaneous tumors in mice. Fourteenth report (summary) 92 

, Hohnes, Harriet F., and Wells, H. Gideon, The occurrence of squa- 
mous cell carcinoma in mice (summary) 180 

Smith, George Gilbert, Radium in cancer of the bladder (summary) 100 

Squamous cell carcinoma in mice. The occurrence of (summary) 180 

Strong, Wendell M., Is cancer mortality increasing? 251 

Sugiura, Kanematsu, Noyes, Helen Miller, and Falk, K. George, The influence 
upon the growth of transplanted Flexner-Jobling rat carcinoma of 
hydrogen ions and of various salts in different concentrations 285 

Sugiura, K., and Benedict, Stanley, R., The influence of certain diets upon 
tumor susceptibility and growth in albino rats (summary) 100 

Teratomas and their relation to age 261 

Theis, Ruth C, The protein content of the whole blood and plasma in cancer. 127 

Thyroid, Malignant tumors of the (summary) 194 

Tissue, islet. An atypical adenoma of the pancreas originating in 277 

Tumor age and tumor incidence 197 

, biology, A phase of (simunary) 96 

, growth. Parabiosis and 121 

, growth. Pregnancy and 245 

, metastases. The influence of heredity in determining 139 

, metastasis. The influence of heredity in determining (summary) 182 

susceptibility and growth in albino rats, The influence of certain diets 

upon (smnmary) 100 

, transplantable. Factors underlying susceptibility to a, in mice 

(smnmary) 106 

Tumors, Experimental production of (summary) 196 

in mice. Further investigations on the origin of 197 

, Malignant, of the thyroid (summary) 194 

, Primary spontaneous, in the kidney and adrenal of mice 305 

, primary spontaneous, of the ovary in mice (summary) 92 

, sarcomatous, Inoculation of, into negro fowls with special reference 

to the significance of chromatophores 31 

, spontaneous, Studies on the incidence and inheritability of, 

in mice 57, 139, 305 

, spontaneous, Studies in the incidence and inheritability of, in mice 

(summary) 182 



386 INDEX 

Tumors, transplantable, Effect of blood from immune anhnals upon 25 

, transplanted animal, The effect of combined heat and radiation upon 

(summary) 101 

, transplanted spontaneous. Effect of a reduction of lyii^)hocytes on the 

growth rate of, in mice 1 

Wells, H. Gideon, Slye, Maud, and Holmes, Harriet F., Primary spontaneous 
squamous cell carcinomas in mice. Studies on the incidence and inherita- 
bUity of spontaneous tumors in mice. Fifteenth communication 67 

, Slye, Maud, and Holmes, Harriet F., Primary spontaneous tumors in 

the kidney and adrenal of mice. Studies on the incidence and inherita- 
bility of spontaneous tumors in mice. Seventeenth communication 306 

, Slye, Maud, and Holmes, Harriet F., Primary spontaneous tumors of 

the ovary in mice. Studies in the incidence and inheritability of spon- 
taneous tumors in mice. Fourteenth report (summary) 02 

— , Slye, Maud, and Holmes, Harriet F., The occurrence of squamous-cell 
carcinoma in mice (summary) 180 

Wilson, Louis B., Malignant tumors of the thjnroid (summary) 104 

Wood, F. C, Experimental production of tumors (summary) 106 

, and Prime, Frederick, Further studies in radiation dosage (summary) . . 177 

X-ray, spectra, High frequency 170 






H 



i/Z 



■> 




I •